Health anxiety: digital interventions appear helpful, but the evidence remains weak

Digital interventions appear to reduce health anxiety in adults. But a systematic review highlights an important limitation: results vary widely across trials, and the certainty of the evidence remains low. The benefit cannot be generalized to just any mental health app.

Illustration of a computer-based consultation about anxious health-related concerns.
MindsetEditorial illustration, not study data.

A review focused on health anxiety

Health anxiety here refers to anxious concerns about illness. The review assesses digital behavioral interventions, including internet-delivered cognitive behavioral therapies. It therefore cannot be used to judge all well-being apps indiscriminately.

The authors searched six databases for studies up to November 1, 2025. They included 16 studies involving 2,581 adults: nine randomized trials, six quasi-experimental studies and one cohort study.

The main meta-analysis, however, draws on a smaller set: six randomized trials involving 585 participants. This distinction matters, because the review’s total sample size is not the one supporting the main estimate of benefit.

A reduction in symptoms, with an uncertain magnitude

Compared with control conditions, digital interventions were associated with a statistically significant reduction in health anxiety. The standardized difference, expressed as Hedges’ g, was −0.77, with a 95% confidence interval from −1.25 to −0.30. This figure is neither a percentage improvement nor an individual probability of recovery.

Results differed substantially between trials, with heterogeneity of 86.7%. The benefit remained statistically significant when each trial was removed in turn, but this analysis does not resolve the differences between studies or their methodological limitations.

No established general psychological benefit

The pooled analyses showed no statistically significant effect on general anxiety, depressive symptoms, anxiety sensitivity or illness-related attitudes. This does not demonstrate an absence of effect: the results remain inconclusive and highly heterogeneous.

Of the nine randomized trials, four had a low overall risk of bias, two raised some concerns and three had a high risk. The authors rated the certainty of the evidence as low for health anxiety and very low for secondary outcomes.

What you can reasonably take away

This review supports the potential value of digital behavioral interventions targeting health anxiety. It is not enough to identify the best program, who benefits most or how long the benefit lasts.

The available abstract does not specify the duration of the interventions or provide details of the comparators. It also does not allow an assessment of adverse effects. The authors call for more robust trials with appropriate active comparators and longer follow-up.

Limitations to keep in mind

Only the PubMed abstract was consulted, not the full text. High heterogeneity, risks of bias and low certainty of the evidence limit the conclusion. Funding and conflicts of interest were not checked.

What we can take away

There is a favorable signal for health anxiety, but it justifies neither a promise of individual results nor a generalization to all apps.

Sources and provenance

Partager cet article

X f in

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