3 - (SYM 4) Is Dosage of App-based Meditation Practice Associated with Changes in Psychological Distress? It Depends How You Ask

Background: Rates of anxiety and depressive disorders have escalated in recent years, driven in part by psychological distress related to the COVID-19 pandemic and other global emergencies. Over this same period, there has been substantial popular and scientific interest in meditation apps as a tool for helping decrease psychological distress at scale. However, very little is known about the dosage of app-based meditation practice that is necessary to achieve treatment effects. This study used data from a recently completed clinical trial testing a meditation app in a sample of predominantly distressed public-school employees (80% with elevated anxiety and/or depression symptoms) to evaluate dose-response associations across a range of operationalizations of dosage and methods for modeling outcomes.

Method: Public school employees (mean age = 42.58 years, SD = 10.66, 87% female, 86% non-Hispanic white) were randomly assigned to use the Healthy Minds Program (HMP) meditation app (n = 344) or to a waitlist control (n = 318). Participants completed a composite measure of psychological distress (i.e., anxiety symptoms, depression symptoms, and perceived stress) at baseline, weekly during the 4-week intervention, and at 3-month follow-up. To thoroughly examine dose-response associations, data visualization and a series of 40 regression models were employed. Latent class analysis was also considered as an alternative operationalization of dosage.

Results: The magnitude of the effect of meditation dosage (and its accompanying significance) varied across dosage operationalizations and methods for modeling psychological distress. Among those assigned to the HMP app, when distress was operationalized as random slopes extracted from multilevel models, higher dosage was generally associated with decreased distress. However, those assigned to the app who did not use it had greater decreases in distress than those who did use it. Several models showed no association.

Conclusions: Results illustrate that answers to questions regarding the clinical effects of meditation apps inherently depend on dosage operationalization and modeling approaches. Careful consideration of these issues is critical, particularly when using specific dosage operationalizations to personalize or tailor app-based intervention delivery. Real-world implications of operationalizations and models considered here, as well as the potential of future work considering more complex operationalizations based on dosage patterns, will be discussed.