In a randomized clinical trial of usual care plus the experimental smartphone-based intervention known as LiveWell vs. usual care alone, participants in the smartphone group who were categorized as low-risk or in asymptomatic recovery at baseline also showed reduced manic symptom severity.
The results suggest that “apps for individuals with bipolar disorder will likely be useful for some people in managing medication use, sleep duration, routine, and monitoring for and managing signs and symptoms” of the disorder, coinvestigator Evan H. Goulding, MD, PhD, assistant professor of psychiatry and behavioral sciences, Northwestern University, Chicago, told this news organization.
Use of the app may also “lead to decreased recurrence of mood episodes, impact overall depressive and manic symptom levels, and improve some aspects of quality of life,” Dr. Goulding added.
The findings were published online in JAMA Psychiatry.
The researchers randomly assigned 205 patients with BD to receive either usual care (n = 81; 56% women; mean age, 39 years) or usual care plus the smartphone-based self-management intervention LiveWell (n = 124; 65% women; mean age, 43 years) between March 2017 and April 2020. To be included, participants could not be experiencing a current mood episode or suicidal ideation.
The smartphone intervention included a daily check-in to monitor medication adherence, sleep, and wellness levels; coach visits to support adherence to the app; six phone calls over 16 weeks; and support from mental health professionals whenever needed. Participants in this group were asked to engage their mental health providers in the intervention as well.
Each participant in the control group had a visit with a coach who facilitated self-management support.
Investigators assessed all participants every 8 weeks until week 48 to gather information on mood symptoms and severity over the past 2 weeks and on quality of life.
The patients were also stratified into high- and low-risk relapse groups. The low-risk group was in asymptomatic recovery, meaning that they experienced two or fewer moderate symptoms of mania or depression in the previous 8 weeks. In addition, they had no moderate symptoms of mania or depression at study enrollment.
Patients in the high-risk group were recovering from an episode of mania or depression. They also had two or fewer moderate symptoms, but for 8 weeks or less.
Low-risk group fares better
Results showed that the smartphone intervention was significantly associated with a reduction in depressive symptoms vs. usual care (P = .02), as well as improvement in one aspect of the World Health Organization Quality of Life Assessment that measures social relationships (P = .02).
When the investigators stratified participants into risk groups, they found that for those in the low-risk group the smartphone-based intervention was associated with lower episode-relapse rates, lower mean percentage time symptomatic, and decreased manic symptom severity.
Mean estimated relapse rates by 48 weeks for the low-risk group were 12% for those in the intervention group and 37.2% for those in the control group. No differences were noted for the high-risk group.
Low-risk patients in the intervention group also had lower mean percentage-time symptomatic (17.9%) than those in the control group (26.1%) (Cohen d = .31).
“Our results are consistent with literature emphasizing the identification and facilitation of management plans for early warning signs of mood episodes and using these plans as an important self-management technique for avoiding relapse,” Dr. Goulding said.
Study limitations included low engagement by mental health professionals and low data generalizability to other populations, as the sample was mostly White (84% of the app group and 81% of the control group).
“There is a fairly large literature on risk factors, longitudinal trajectories, and stages of diseases that suggest we should already be able to predict relapse risk for individuals,” Dr. Goulding said.
“However, moving from overall risk to individual risk is trickier and will require larger datasets with longer follow-up to better understand what types of help should be delivered when and to whom,” he added.