Wearables and weight loss: why more steps may not move the scale

Wearables and apps can increase activity, but weight loss is usually modest. Study results depend on the support provided and how long it continues.

日本語で読む（Japanese version）

Your wrist shows a step count. Your phone adds a leaderboard, reminders, and a graph. When choosing a tool to track activity, it is easy to assume that more information will mean more weight loss. Before buying, or deciding whether your current device is working, separate what the screen can display from what a study actually found.

In a randomized controlled trial of 109 people with an average age of 47, the group using an app became more physically active. Yet both groups lost 2.2 kg, with no statistically significant difference between them.[11] Activity increased, but weight loss did not differ between the groups. Understanding both results is more useful than treating a rising step count as a promise about the scale.

A wearable device is different from a weight-loss program

Here, a wearable is a device worn on the body to record activity. Researchers investigate what changes when people receive an intervention involving that device, compared with people receiving a different intervention. Some programs focus on activity tracking. Others add advice about food and exercise. Similar-looking watches can therefore be part of quite different treatments, and the appearance of the hardware tells you little about the full program being tested.[1][3]

A randomized controlled trial assigns participants to groups by chance and compares their results. Antoun and colleagues used meta-analysis to combine the results of trials involving apps and accompanying support. These distinctions matter when a headline says that an app helped people lose weight. If the studies also supplied coaching or an activity tracker, the combined result cannot be read as the effect of the app alone. Antoun and colleagues explicitly examined accompanying support in their analysis of 34 studies.[3]

Imagine someone buying a watch after reading that a digital intervention produced weight loss. Now imagine someone using the same kind of record to review food and activity, adjust a plan, and receive feedback from another person. Those people have access to different forms of support. To bring a research finding into everyday life, ask what happened after the measurement, as well as what the device measured. The tested intervention may involve a whole sequence of actions beyond putting something on a wrist.[3][12]

Automatic recording can make a tool easier to use. It does not, by itself, explain how its information will influence tomorrow’s behavior. Looking at the steps from your journey home is one possibility. Looking back at the route and time that made walking possible, then planning another opportunity, is another. The latter is an everyday application of elements used in the research: self-monitoring, goals, action planning, and feedback.[12] Charging the watch does not fill in tomorrow’s schedule.

This also helps distinguish related questions. The accuracy of a home body-composition scale belongs to the feature on body-fat measurement. How detailed a food record should be belongs to the feature on self-monitoring. Here, the central issue is whether adding a digital tool changes activity and weight. A device’s ability to generate a number, the use made of that number, and the outcome being assessed need separate attention. Otherwise, a measurement feature can quietly become a claim about weight loss without the relevant comparison ever being examined.

Steps often increase, while weight changes are smaller

Ferguson and colleagues’ 2022 study was an umbrella review: a review of systematic reviews and meta-analyses. It included 39 reviews and meta-analyses, reporting results from 163,992 participants across a broad range of ages, including both healthy people and people with medical conditions. Overall, activity trackers improved physical activity. The authors translated the findings into approximately 1,800 extra steps per day, 40 more minutes of walking per day, and a reduction of about 1 kg in body weight.[7]

The authors described the benefits as clinically important and sustained over time. They concluded that the studies provided sufficient evidence to recommend activity trackers. That positive conclusion should remain part of the picture. Equally, effects on blood pressure, cholesterol, glycated hemoglobin, quality of life, and pain were generally small and often not statistically significant. Activity benefits did not mean that every physiological or quality-of-life measure improved to the same extent.[7]

The figures for steps and walking time are approximate translations of pooled findings. They are not instructions for every reader to add exactly that amount to an existing routine. Nor do they promise that someone who adds those steps will lose the reported amount of weight. Their value is in showing the scale of the outcomes in the evidence, while keeping activity and weight in their own units. A research average is useful context for a walking plan, but it does not supply a personal prescription.[7]

Vosadi and colleagues focused on adults with overweight or obesity. Their meta-analysis included 26 studies and 2,373 participants using wearables or smartphone-based interventions. The difference in weight between intervention and comparison groups was −1.61 kg. Steps and moderate-to-vigorous physical activity also improved significantly. However, differences in blood pressure and resting heart rate were not significant. Improvement in one outcome was not evidence that all the health indicators on a dashboard would improve.[6]

Tang and colleagues examined healthy adults more specifically. Their analysis included 12 trials and 1,693 participants. The weighted average age was 40.7 years, 64.4% were women, and the average intervention lasted 21.4 weeks. Overall physical activity improved, with a standardized mean difference of 0.449 and a 95% confidence interval of 0.10 to 0.80. In subgroup analyses, however, the trackers showed no clear benefit for physical activity or weight reduction.[1]

Tang’s activity estimate puts different measurements on a common, unitless scale; its interval of 0.10 to 0.80 shows the uncertainty around that estimate. The activity effect of 0.449 is therefore not a number of kilograms, minutes, or steps. Tang and colleagues concluded that trackers might produce modest short-term increases in activity and that further evidence was needed to establish sustained benefits. The overall result and the less clear subgroup results both belong in an account of this study.[1]

A broad umbrella review can reach a favorable conclusion while an analysis limited to healthy adults remains cautious. Different participants, durations, and comparisons answer different questions. There is no need to sort the papers into simple camps for or against wearables. For a reader deciding what to expect, the shared practical lesson is to assess the possibility of increased activity separately from the size and persistence of any weight change.[1][6][7]

How much difference can an app make?

Pujia and colleagues’ 2025 meta-analysis restricted its question to smartphone-only interventions in adults with overweight or obesity who did not have accompanying medical conditions. Studies using additional devices were excluded. It included 11 randomized controlled trials with 1,717 participants, and interventions lasted from 60 days to 12 months. The programs involved monitoring diet and exercise through smartphone apps. This narrower selection helps identify what can be said about that particular type of intervention.[5]

At 4–6 months, the weight effect was a standardized mean difference of −0.33, with a 95% confidence interval of −0.48 to −0.17. The effect was statistically significant but small. It is a standardized effect, not a change measured in kilograms. Body-fat percentage also improved at some assessment points, while waist circumference did not show a significant effect. The authors highlighted limited effectiveness beyond 6 months, a tendency toward weight regain, and problems with continued engagement.[5]

Another 2025 analysis, by Li and colleagues, combined 29 studies in adults with overweight or obesity. The difference in weight for app-based interventions versus comparison groups was −1.45 kg, with a 95% confidence interval of −2.01 to −0.89 kg. Significant differences were also reported for waist circumference, at −1.98 cm, and fat mass, at −1.32 kg. Its selection of studies differed from the smartphone-only analysis, so the two results describe different bodies of evidence.[12]

Li and colleagues also reported a reduction in diastolic blood pressure, the pressure when the heart relaxes between beats. It was 1.76 mm Hg lower than in comparison groups, with a 95% confidence interval of −3.47 to −0.04 mm Hg. A difference was found in glycated hemoglobin, a blood-glucose indicator, as well. Other assessed outcomes did not show significant effects. The review looked beyond weight to behavioral and metabolic outcomes, and its favorable results were specific to the measures that changed.[12]

Put the waist findings side by side: one analysis found no significant effect, while another found a significant reduction. This illustrates why an app’s effectiveness cannot be captured in a single universal figure. Participants, accompanying support, and assessment times can differ. Rather than assuming that one paper cancels out the other, ask which intervention is closer to the way you expect to use an app. The narrower eligibility criteria of the smartphone-only review are part of the explanation for why its answer has a different scope.[5][12]

Yen and colleagues’ 2022 meta-analysis addressed apps aimed at changing physical activity. Across 12 trials, it reported small-to-moderate effects on weight, body mass index, and body-fat percentage. Subgroup analyses found larger effects in interventions with reminders, self-reporting, and a health coach. Studies involving older participants or longer interventions also had larger effects. Weight effects were larger among participants with a disease or disability, and body-mass-index effects were larger among participants with overweight or obesity.[2]

These findings make the content of support worth examining, but they do not establish a simple shopping rule. Antoun and colleagues found no association between the type or number of app features and weight loss across their 34-study analysis. They did find significant weight differences when apps were combined with intensive behavioral coaching or human feedback. A list of notifications, graphs, and rankings is therefore less informative than an explanation of how the program uses those features to support actions.[3]

For app-based interventions overall, Antoun and colleagues reported a between-group weight difference of −1.99 kg at 3 months, with a 95% confidence interval of −2.19 to −1.79 kg. At 6 months, the difference was −2.80 kg, with a 95% confidence interval of −3.03 to −2.56 kg. Results varied substantially between studies: the heterogeneity measures were 81% and 91%, respectively. The confidence interval around the average and the variation between individual studies convey different information; both matter when interpreting the pooled figure.[3]

Numbers: wearables, apps, and weight

Approximately 1,800 extra steps per day: the activity increase expressed in everyday units in the broad umbrella review.[7]

About 1 kg: the approximate body-weight reduction reported in that same review, alongside its activity findings.[7]

−1.61 kg: the between-group weight difference across 26 studies in adults with overweight or obesity.[6]

−1.45 kg: the weight difference across 29 studies in adults with overweight or obesity, with a 95% confidence interval of −2.01 to −0.89 kg.[12]

−2.2 kg and −2.2 kg: weight loss in the two groups of the 109-person trial. Activity differed, but the weight-loss difference was not significant.[11]