January is peak gym time — the month when fitness seekers commit to New Year’s resolutions. By February, however, those goals are often forgotten. That busting of best intentions begs the question: how much exercise do people really need, and what’s the ideal way to get it? Recent research fortunately has some welcome advice for the time-stressed in 2026.
Why is exercise good for you? Scientists are finding answers in our cells
Existing guidelines from most national and global health organizations call for at least 150–300 minutes of moderate physical activity each week, or 75–150 minutes of vigorous activity, for healthy adults, sometimes alongside activities to strengthen muscle and bone. Although those guidelines remain good goals to aim for, newer studies suggest that meaningful health benefits emerge with much less exercise.
Researchers are getting a clearer understanding of the bare minimum amount of exercise needed for health gains thanks to data from wearable devices. These can provide more-reliable measurements than do self-reported data, which form much of the basis for current guidelines. By incorporating wearables into study design, researchers can collect accurate data on physical activity minute by minute, says I-Min Lee, an epidemiologist at the Harvard T.H. Chan School of Public Health in Boston, Massachusetts. “And this is when we start to see that even low levels of physical activity are helpful,” she says.
These data are redefining what counts as physical activity and could push future recommendations towards lower targets. But a potential shift in guidelines needs to be considered carefully in terms of the message it might send, says Emmanuel Stamatakis, a researcher specializing in physical activity and population health at the University of Sydney, Australia. No one wants to suggest that people shouldn’t strive to move more, especially when an estimated 31% of people worldwide don’t meet existing recommendations1, and physical inactivity contributes to health problems such as obesity and heart disease.
“We’re in a bit of an awkward situation because if we are to take the wearable-based evidence at face value, we will have to release guidelines that will be recommending lower amounts” of minimum physical activity, says Stamatakis.
Another way in which wearables are changing thinking is by helping to quantify daily inactivity, which might be just as important as how much a person exercises. Some guidelines are already incorporating caps on sedentary time. “Especially if you work from home, you can do very little exercise in a day. It’s quite scary,” says Carol Maher, an exercise researcher at the University of South Australia in Adelaide.
The more, the better — up to a point
Today’s exercise recommendations are born from large epidemiological studies that compare disease and death rates between those who are more and less physically active. Whereas previous guidelines had focused on enhancing athletic performance and relied mostly on studies in young and fit male medical students, studies beginning in the 1980s followed larger groups of people over many years and included women and older people.
Such studies have consistently shown that physical activity offers strong protection against cardiovascular disease, reduces the risk of several types of cancer and lowers the risk of death from any cause, as well as having mental-health benefits. Although observational studies can be limited by potential bias — those who exercise might be healthier to start with — researchers go to great lengths to mitigate this problem. Studies tend to control for variables such as age, smoking status, alcohol consumption and body weight, and adjust for family history of cardiovascular disease and cancer, as well as postmenopausal hormone use, sleep duration and other factors. Some studies exclude participants diagnosed with cardiovascular disease or cancer and those who do not engage in any physical activity. Some also exclude the first years of observation, so that participants with as-yet-undiagnosed diseases can be identified over time. The biases alone cannot explain the findings, says Lee.
... continue reading