Scaling a Brief Digital Well-Being Intervention (the Big Joy Project) and Sociodemographic Moderators: Single-Group Pre-Post Study.

Abstract

Emotional well-being interventions lead to better mental and physical health. However, most of these interventions have been tested on relatively homogeneous samples, with few interventions large enough to examine whether key sociodemographic factors impact outcomes. In addition, barriers to engagement include access and high participant burden. We developed a brief web-based intervention to address these barriers and tested the effects across sociodemographic groups. The study aims to examine the effectiveness of a brief, low-burden digital well-being intervention in improving emotional well-being and health-related outcomes across a diverse global sample. It investigates how key sociodemographic factors, such as age, sex, race and ethnicity, education, financial strain, and subjective social status, moderate intervention effects. The goal is to identify which groups benefit most, informing the scalability and public health potential of digital well-being interventions. We conducted a single-group pre-post study spanning from 2022 to 2024 using a web-based, multicomponent, week-long well-being intervention requiring 5 to 10 minutes of daily activity (the Big Joy Project). Using a global convenience sample recruited via open web-based enrollment, we assessed pre-post changes in emotional well-being, positive emotions, happiness agency, perceived stress, self-reported health, and sleep quality. At baseline, participants also reported sociodemographic characteristics. We used mixed-effects linear models to examine pre-post changes in the outcomes and sociodemographic moderators. The sample (N=17,598) consisted of individuals from 169 countries and territories, with broad representation across sociodemographic groups; however, the sample was predominantly White, female, and had at least a high school or college education. Following the intervention, participants showed significant within-subjects effect size increases in emotional well-being (d_z=0.48; P<.001), positive emotions (d_z=0.45; P<.001), and happiness agency (d_z=0.44; P<.001). Furthermore, participants showed a decrease in perceived stress (d_z=-0.35; P<.001) and an increase in self-reported health (d_z=0.07; P<.001) and sleep quality (d_z=0.15; P<.001). There was a clear dose-response pattern across outcomes: participants who engaged in more daily practices showed greater improvements. There was a strong pattern of social disadvantage moderating these effects, with groups experiencing greater social disadvantage showing larger benefits across most outcomes. For example, those with lower education, greater financial strain, or lower subjective social status and those identifying as individuals from racial or ethnic minority groups (Black or Hispanic) all showed larger improvements across well-being outcomes. Furthermore, younger people had greater increases in emotional well-being and greater decreases in perceived stress compared to older people. A brief, low-intensity intervention showed meaningful improvements in well-being and stress, comparable to those seen in longer, more intensive digital well-being interventions. Sociodemographic groups that are at higher risk of poor mental health benefited more from the intervention, highlighting its potential for scalable public health impact. Testing this intervention with a randomized controlled trial design will be important.