The best health apps pair exercise with self-management coaching

Apps, wearables and telehealth are increasingly pitched at people with long-term conditions who need to move more, but the offerings vary wildly. Some are little more than a web platform, while others add phone calls, devices or full exercise programmes. Until now nobody had unpicked which ingredients actually drive the increase in activity, which leaves patients and clinicians choosing between products on marketing rather than evidence. This review set out to isolate the effect of each component and each delivery mode, so future interventions can be built from the parts that genuinely work.

Researchers pooled randomised controlled trials of digital health interventions targeting physical activity in people with a chronic condition or several at once. They searched four major databases up to March 14, 2024 and included 137 trials with 18,443 participants. Rather than a simple head-to-head pooling, they ran a component network meta-analysis, a technique that estimates the contribution of individual ingredients such as exercise programmes, self-management support and different technologies. They also rated how much confidence each comparison deserves using an established certainty framework, and probed the findings with subgroup and sensitivity analyses.

Exercise combined with self-management support beat both usual care and minimal self-management for raising physical activity, and self-management alone also outperformed both comparators. On delivery, mixed technologies such as apps paired with phone calls or devices came out strongest, followed by web and app platforms, with phone calls alone showing a smaller and less certain effect. People with type 2 diabetes showed the greatest increase in activity, while those with musculoskeletal conditions improved least. Certainty of the evidence was moderate for most comparisons, but low for phone calls.

Digital health tools have been tested in hundreds of trials, but results have been inconsistent, partly because interventions bundle very different components under a single label. This analysis moves the field forward by ranking ingredients rather than whole products, and it supports the intuition that structure plus support beats information alone. Open questions remain: why muscle and joint conditions responded least, whether the gains persist after a programme ends, and how much the human element behind mixed technologies matters compared with the technology itself.

If you or someone you know is managing a long-term condition and wants to be more active, this is a reason to favour digital tools that offer an actual exercise programme alongside self-management support, rather than tracking or information alone. It also suggests a blend of channels, an app plus calls or a device, tends to deliver more than any single one. These are averages across many conditions, so individual results will differ, and the gains were modest rather than dramatic.