People taking GLP-1 weight-loss medications such as semaglutide and liraglutide ended up moving less in their daily lives, not more, according to a new analysis of Fitbit data from 753 participants in the National Institutes of Health’s All of Us Research Program. The findings, presented at ENDO 2026, challenge a widely held assumption that shedding pounds on these drugs would naturally lead to higher physical activity. Instead, the data point to a potential blind spot in how millions of users and their doctors think about the full metabolic picture of GLP-1 treatment.
Why a Drop in Daily Movement Changes the GLP-1 Calculus
The promise of GLP-1 receptor agonist drugs has always extended beyond the number on the scale. Weight loss is supposed to make movement easier, reduce joint pain, and encourage a more active lifestyle. That logic made intuitive sense to clinicians and patients alike. The new wearable-data analysis flips that expectation: researchers found that people who started these medications recorded fewer fairly active minutes and took fewer daily steps over time, based on Fitbit-tracked activity in 753 participants with sufficient device data.
The tension here is practical. If GLP-1 drugs suppress appetite and reduce caloric intake but also dampen the low-grade physical activity that burns calories throughout the day, some of the expected cardiometabolic benefit could be lost. Non-exercise activity thermogenesis, the energy spent on everything from fidgeting to walking to the mailbox, accounts for a meaningful share of daily calorie expenditure. A drug-driven reduction in that background movement, especially one that scales with dose or treatment duration, would represent a real trade-off that prescribers and patients need to account for.
There is also a psychological dimension. Many patients view GLP-1 treatment as a chance to “reset” their lifestyle, imagining that once the weight starts coming off, they will feel more motivated to work out. If, in reality, the medication makes them feel slightly more fatigued or less inclined to move, that expectation gap could lead to frustration and disengagement from other healthy behaviors. The ENDO 2026 data do not prove cause and effect, but they raise the possibility that weight loss achieved pharmacologically may not automatically unlock a more active daily routine.
Fitbit Data, the All of Us Cohort, and What the Numbers Show
The study drew on the All of Us Research Program’s wearables dataset, which collects Fitbit-derived daily summaries including steps, sedentary minutes, light active minutes, fairly active minutes, very active minutes, and sleep metrics. A peer-reviewed methods paper describes how Fitbit intensity categories are defined by metabolic-equivalent thresholds, giving researchers a standardized way to classify movement levels across thousands of participants. The dataset itself spans millions of observations across a large participant pool, making it one of the most extensive real-world sources for tracking how daily behavior shifts during drug treatment.
The 753-person analytic cohort was drawn from All of Us participants who had both GLP-1 prescription records and enough consecutive Fitbit wear days to produce reliable before-and-after comparisons. The key finding, that fairly active minutes declined after treatment initiation, ran counter to what the research team anticipated. Lighter people generally move more, so the working hypothesis had been that weight loss would correlate with increased activity. The Fitbit data told a different story.
Separate randomized trial evidence adds an important wrinkle. A post-treatment analysis of a placebo-controlled trial comparing exercise, liraglutide, and the combination of both found that accelerometer-measured sedentary time was similar between groups after treatment ended. That trial showed combination therapy produced the best long-term weight maintenance, but the sedentary-time finding suggests that structured exercise programs do not automatically translate into less sitting. The ENDO 2026 results extend that observation into real-world, uncontrolled settings where most patients do not have a supervised exercise protocol.
A complementary 12-week wearable study examined behavioral and physiologic responses during GLP-1 receptor agonist treatment, tracking heart rate and activity shifts in the early weeks after drug initiation. In that work, published in a digital health journal, investigators used consumer wearables to monitor early changes in movement and cardiovascular strain, showing that measurable shifts in activity can emerge within weeks of starting therapy. Together, these strands of evidence suggest that GLP-1–related activity changes are not a late, secondary effect but part of the early adaptation to treatment.
Open Questions About Dose, Duration, and Lost Calories
Several gaps in the evidence remain. The ENDO 2026 presentation did not include published pre-and-post step counts, exact changes in MET-minutes, or p-values for the activity decline. Without those granular numbers, it is difficult to know whether the drop in fairly active minutes is clinically meaningful or a modest statistical signal. The analytic cohort of 753 is large for a wearable study but still a fraction of the millions now using these drugs, and the population that owns and consistently wears a Fitbit skews younger and more health-conscious than the general GLP-1 user base.
The most pressing unanswered question is mechanistic. GLP-1 receptor agonists act on brain regions that regulate appetite, reward, and energy balance. Whether the drugs directly suppress the neural drive to move, or whether reduced caloric intake simply leaves people with less energy for spontaneous activity, has not been established. If the effect is dose-dependent, patients on higher doses of semaglutide or tirzepatide could see larger activity declines, a possibility that no published study has yet quantified.
Calorie intake data is also absent from the wearable analyses. Without knowing how much less participants are eating, it is impossible to calculate whether the reduction in daily movement fully offsets, partially offsets, or barely dents the caloric deficit created by appetite suppression. A recent analysis of GLP-1–related behavior change, published in a clinical medicine journal, highlighted how hard it is to disentangle medication effects from shifting diet and lifestyle in real-world settings. That work, available as an open-access study, underscores the need for integrated datasets that pair prescriptions with both objective activity tracking and detailed nutritional records.
Duration is another unknown. The All of Us analysis compared activity before and after treatment initiation, but it is not yet clear whether the observed decline plateaus, worsens, or gradually recovers as patients adapt to the medication. Longitudinal follow-up over years, rather than months, will be needed to understand whether a temporary dip in movement is a tolerable side effect or a persistent pattern that blunts long-term cardiometabolic gains.
What Clinicians and Patients Can Do Now
While the science catches up, the practical implications are straightforward. Clinicians prescribing GLP-1 drugs may need to be more explicit about protecting daily movement, not just encouraging formal workouts. That could mean setting step-count targets, reviewing wearable summaries during follow-up visits, or flagging unexplained drops in fairly active minutes as a reason to adjust counseling.
Patients, meanwhile, can treat the medication as one tool among many rather than a standalone fix. Using a fitness tracker to monitor steps and active minutes, building short walking breaks into the workday, and planning light-intensity movement on days when appetite is particularly low are all ways to guard against an unnoticed slide into sedentariness. Even modest increases in non-exercise activity-taking stairs, doing household chores, pacing during phone calls-can help preserve the calorie-burning background hum that GLP-1 therapy may otherwise dampen.
The emerging message from ENDO 2026 and related research is not that GLP-1 drugs are ineffective or inherently harmful, but that their benefits are intertwined with behavior in more complex ways than early marketing narratives suggested. Weight loss alone is not the whole story. For patients and prescribers hoping to maximize cardiometabolic health, paying attention to what happens between the gym sessions-every step, every fidget, every walk to the mailbox-may prove just as important as the number on the prescription label.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
