Outcomes of a Gratitude Practice in an Online Community of Caring.

Objectives: CaringBridge (CB) is a web-based social network where people share information, enlist support, and access resources following a difficult diagnosis; it can also be used to disseminate supportive self-care tools, such as a gratitude practice, for its users. Gratitude practices are shown to reduce stress and fear, improve sleep, and increase positive emotions and overall well-being. The purpose of this article was to report the findings of a brief gratitude intervention delivered to CB users. Design, setting/location, subjects: This is a nonrandomized, prospective, pre- and post-evaluation study in an online community. Inclusion criteria were adults 18 years or older, English literate, willingness to participate in a mind-body practice, and active users of CB: patient, caregiver, or visitor to a site.

Interventions: Participants were engaged in a daily, 21-day brief gratitude practice and were given weekly automated reminders to do their practice.

Outcome Measures: Outcomes included perceived stress, gratitude, social connectedness, and social assurance scales. Paired t tests were used to assess changes in outcomes; multivariate regression models were used to assess the relationship between the frequency of gratitude practice and change in outcomes.

Results: Follow-up data were collected from 882/1598 participants, and nearly 70% self-reported engaging in the gratitude practice five or more days/week. Participants reported statistically significant improvement in all outcomes with small standardized effect sizes for gratitude (0.39), social connectedness (0.24), and social assurance (0.10). Changes in perceived stress (-0.73) were larger in magnitude and increased with more frequent practice.

Conclusions: Among this online community, there was a high level of engagement with a brief gratitude practice, and improvements in stress, gratitude, and social support were observed. This design did not control for changes in outcomes that may be due to time trends, placebo or contextual effects, regression to the mean, or selection bias.