Type 2 diabetes: app use associated with better outcomes, without evidence of causality

An app can support diabetes management, but is using it more enough to improve outcomes? In a secondary analysis of a trial involving adults with type 2 diabetes, greater use was associated with better self-care behaviors and lower glycated hemoglobin at six months. These associations do not demonstrate that the app causes these improvements.

Illustration of a smartphone app for tracking diabetes.
Gadgets & TechEditorial illustration, not study data.

An analysis focused on participants with access to the app

The Living Healthier with Diabetes trial, conducted in Texas between 2020 and 2022, included 189 participants divided into three groups: a remote education and support program, a smartphone app-based program, and a combined intervention.

This secondary analysis covers only the 129 participants in the two groups with access to the app. It therefore does not directly compare outcomes with and without the app. The researchers examined the relationships between baseline attitudes toward technology, use of the tool, and outcomes at three and six months.

Use was assessed using an index combining measures including days of use per week, minutes per day, and overall use over three months. Another index measured the use of ten app features.

Associations varied by outcome and time point

Participants who initially had a more favorable attitude toward technology used the app more at three and six months. However, this attitude was not associated with the index measuring use of the different features.

Greater use was associated with better diabetes self-care behaviors at both time points. For glycated hemoglobin, or HbA1c, an association with a lower value appeared at six months, but not at three months.

The analyses accounted for intervention group, age, gender, financial situation in the previous month, and baseline values of the outcomes studied. These adjustments reduce some alternative explanations without eliminating them all.

What these findings tell us

Although the data come from a randomized trial, the amount of app use was not itself randomly assigned. People who use it more may also differ in other characteristics related to their diabetes management. This analysis cannot isolate the effect of app use itself.

The finding concerns an app integrated into a structured education and support program. It does not demonstrate that any commercial app would produce the same effects. The abstract also does not provide an absolute difference in HbA1c between clearly defined usage groups that would make it easy to assess the magnitude of the clinical benefit.

Limitations to keep in mind

Only the PubMed abstract was consulted, not the full text. This is a secondary analysis, limited to participants with access to the app, with incomplete follow-up data. Causality and the clinical significance of the association remain uncertain. Funding and conflicts of interest have not been verified in the dossier.

What we can take away

Engagement with a diabetes support app is associated with some favorable outcomes. This study demonstrates neither that using the tool more directly improves blood glucose nor that all apps are equally effective.

Sources and provenance

Partager cet article

X f in

An error or a source to report? Contact the editorial team.