The study was very<\/a> widely<\/a> covered<\/a> in the media and extensively<\/a> discussed<\/a> on social media. The main message, as highlighted in the accompanying press release from Harvard<\/a>, where the researchers are based, is that doing a variety of different types of exercise is healthier than doing just one thing, regardless of how much you do. But commenters were also fascinated by a series of colorful graphs plotting dose-response curves for individual types of exercise, purporting to show that walking is the best, jogging is good but not too much, swimming is no good at all, and so on. And more generally, the results even hint that beyond a relatively modest amount of weekly exercise, doing more of anything doesn\u2019t help you.<\/p>\n It\u2019s worth taking a closer look at these results (the paper is free to read online<\/a>), because they convey some important lessons about what exercise does for us, what it doesn\u2019t do, and why even massive, long-running studies can lead us astray if we don\u2019t interpret them carefully. I\u2019ll take a look at three key claims: that too much exercise is bad; that certain types of exercise are uniquely bad; and that doing a variety of exercises is best.<\/p>\n Is Too Much Exercise Bad?<\/p>\n First, some background: the data in the new paper comes from two well-known cohorts: the Nurses\u2019 Health Study and the Health Professionals Follow-Up Study. Participants periodically filled in questionnaires reporting how much time in a given week they spent doing various activities such as walking, running, cycling, swimming, playing tennis, or lifting weights. That\u2019s the only data the researchers had on physical activity: no details on whether you\u2019re spending that time, say, dog-paddling in the shallow end with your toddler or swimming across lakes at top speed.<\/p>\n The most basic finding is that exercise reduced the risk of premature death. In the graph below, the most important line is \u201call cause mortality,\u201d which expresses your overall risk of premature death compared to someone who doesn\u2019t exercise at all, as a function of how much you exercise:<\/p>\n That all-case mortality line drops sharply and then plateaus at roughly 0.7\u2014that\u2019s a 30 percent reduction in risk\u2014beyond about 20 MET-hours per week of exercise.<\/p>\n Note that this is not hours per week of exercise. A \u201cmetabolic equivalent of task,\u201d or MET, is a measure of how hard exercise is relative to how much energy you burn when you\u2019re just lying on the sofa. For example, they estimate that walking takes 3 METs (triple your resting metabolism), so walking for 2 hours would burn 6 MET-hours of energy. That means 20 MET-hours a week is just under an hour a day of walking. Running, on the other hand, is assumed to take 12 METs, so 20 MET-hours is 100 minutes of running per week.<\/p>\n So the big question here is: Is it pointless to do more than 20 MET-hours per week? Various versions of this question have been debated<\/a> for the last decade or so. One answer: if you enjoy it, knock yourself out. There\u2019s no evidence here that higher levels of exercise are actually bad for you. And intuitively, it makes sense that you eventually reach a point of diminishing returns from exercise.<\/p>\n Personally, though, I\u2019m not convinced that 20 MET-hours per week is a genuine threshold. There are a lot of potential issues with this sort of epidemiological study, where scientists can only observe without asking people to change their behavior. The most important one is the risk of confounding variables: people who exercise a lot tend to be different from people who don\u2019t exercise a lot, and some of those factors\u2014socioeconomic status, smoking, nutrition, and so on\u2014will affect how long people live.<\/p>\n To combat this problem, statisticians \u201cadjust\u201d the results to effectively make it seem as though the subjects were identical in every respect except how much they exercise. For example, in this study, they adjusted the results to equalize diet healthiness based on a score called the Modified Alternative Healthy Eating Index. They have to do this because the most active subjects had a significantly higher diet score than the least active subjects, and we don\u2019t want to mistakenly attribute the longevity benefits of their healthy eating to exercise.<\/p>\n But they also adjust other factors such as BMI, blood pressure, and cholesterol levels. And here we run into a problem of circularity. Exercise lowers BMI, blood pressure, and cholesterol levels, so it\u2019s not surprising that the most active subjects have lower scores in all these categories. But statistically equalizing these parameters effectively erases some of the benefits of exercise! If it appears that exercise has no longevity benefits beyond 20 MET-hours per week, that\u2019s in part because we\u2019re leaving out any beneficial effects of lowering BMI, blood pressure, and cholesterol.<\/p>\n (Side note: yes, many people would argue that exercise doesn\u2019t lower BMI. Let\u2019s leave that argument<\/a> for another day. I will merely note that the average BMIs in this study, by quintile from least to most active, are 26.4, 25.5, 25.0, 24.5, and 23.9.)<\/p>\n What difference does this statistical adjustment make? Here\u2019s what the relative risk of premature death looks like for five roughly equal groups divided up based on how much physical activity they reported. The \u201cadjusted\u201d data corrects for ethnicity, family history of heart disease or cancer, BMI, smoking status, diet quality, blood pressure, cholesterol, and various other factors:<\/p>\n In the adjusted data, the benefits of exercise seem to plateau for the fourth and fifth quintiles. In the unadjusted data, on the other hand, the benefits continue to increase. The gap between the two lines reflects, in part, the degree to which the adjusted results are underestimating the benefits of exercise. This is why I\u2019m not convinced that the benefits of exercise plateau beyond 20 MET-hours per week.<\/p>\n Are Some Types of Exercise Uniquely Bad?<\/p>\n Lots of previous epidemiological studies have shown that exercise, as a whole, promotes longevity. This one is so big that they have the statistical power to break down the results by individual activities. They share the results for nine of these activities in the following graph:<\/p>\n What jumps out is that all the curves are different. Walking produces a big, robust decrease in the risk of premature death (the \u201chazard ratio\u201d) followed by a plateau. Jogging has a U-shaped curve; running has more of a W shape. Wait, what? \u201cJogging,\u201d it turns out, is classified as anything slower than ten minutes per mile, while \u201crunning\u201d is faster than that. This is our first warning sign that something is amiss, since ten-minute miles are a very different form of exercise for a fit 30-year-old and a mostly sedentary 80-year-old. Still, it\u2019s odd that the curves have completely different shapes.<\/p>\n From the researchers\u2019 perspective, the main message here is the unsurprising fact that the curves aren\u2019t straight lines\u2014that is, that the benefits of adding more exercise to your routine depend on how much exercise you\u2019re already doing.<\/p>\n On social media, on the other hand, there\u2019s a lot more commentary on the specific shapes of individual curves. Why, for example, is a little bit of bicycling good, a medium amount bad, and a very large amount good again? You can (and many commenters did) come up with elaborate explanations about how the cardiovascular benefits of cycling interact with the risks of getting sideswiped by a truck to create this odd-shaped curve. But this is clearly nonsense. What we have here is imperfect data fed into complex statistical adjustment algorithms and then chopped into pieces that are too small to yield meaningful insights.<\/p>\n Some of the curves do look reasonable. Walking looks okay (with the caveat, again, that the plateau is probably an artifact of adjusting for things like cholesterol). I like the running curve, which suggests that (counter to earlier claims<\/a>) more is better up to very high amounts. But that\u2019s not how it works: you can\u2019t just pick the outcomes that make sense to you and assume they\u2019re \u201ccorrect\u201d while ignoring the other ones. That\u2019s like the famous quote from ancient Greek physician Galen (which I encountered most recently in David Epstein\u2019s forthcoming book, Inside the Box): \u201cAll who drink of this remedy recover in a short time, except those whom it does not help, who all die.\u201d<\/p>\n Is Variety the Spice of Workout Life?<\/p>\n This brings us to the central message of the study: that doing a variety of different types of physical activity is better for you than an equivalent amount of just one activity. This claim passes the most important hurdle, which is that it makes sense. As the researchers point out, there\u2019s solid evidence that aerobic exercise primarily increases cardiovascular fitness while resistance exercise primarily increases muscular strength. Both strength and cardiovascular fitness have been associated with longevity, so we should assume that variety is good.<\/p>\n To test this claim, the researchers assigned each subject a \u201cvariety score,\u201d which basically involved adding up the number of types of exercise they reported doing consistently. The maximum score was 13. The problem, of course, is that people who have a higher variety score probably do more exercise overall\u2014so if they live longer, you can\u2019t tell whether it\u2019s variety or amount that helped. You need to look at both factors at once.<\/p>\n Here\u2019s a graph showing the risk of premature death for three groups (on the horizontal axis) who had low, medium, or high total amounts of physical activity. Within each group, there\u2019s a data point for people who had low, medium, or high exercise variety:<\/p>\n You can see that, within each group, going from low to high variety mostly seems to be helpful. The effect is clearest for the group doing the most overall exercise, which makes intuitive sense. My guess is that this is a real effect, but in addition to the problems with statistical adjustment that I mentioned above, there\u2019s also a risk here of reverse causation: people who are already healthy are the most likely to play a weekly game of pickleball and jog a few a times a week and hit the gym.<\/p>\n So what should we take from this study, or more generally from the large existing body of epidemiology with all its strengths (big numbers and long timeframes) and weaknesses (messy data and lack of randomization)? A useful thought experiment is to imagine your reaction if the results had been reversed. If this study had suggested that even modest amounts of exercise are bad for you, or that exercise variety shortens your life, I simply wouldn\u2019t have believed it. Since that\u2019s the case, I shouldn\u2019t put too much weight on the findings even where they agree with my preconceptions.<\/p>\n Someday, improved wearable accelerometers are going to give us big data sets where we\u2019ll have a much more accurate picture of how and how much people really exercise. Until then, there\u2019s a parallel body of epidemiological research that skips the messy process of asking people to guess how much of each activity they do. It involves objective measures of fitness. Instead of asking \u201cHow much do you jog?\u201d, they measure your VO2 max<\/a>; instead of asking \u201cHow much do you lift?\u201d, they measure your grip strength<\/a>. Both of these parameters predict longer life, with no evidence of negative effects even at the highest level of fitness. If you\u2019re looking to optimize both your lifespan and your healthspan, the advice remains simple: get as fit and as strong as possible, following whatever plan works for you.<\/p>\n For more Sweat Science, sign up for the email newsletter<\/a> and check out my new book The Explorer\u2019s Gene: Why We Seek Big Challenges, New Flavors, and the Blank Spots on the Map<\/a>.<\/p>\n \u00a0<\/p>\n","protected":false},"excerpt":{"rendered":"Published January 30, 2026 06:26AM People often complain\u2014and rightly so\u2014that the typical exercise science study involves a half-dozen…\n","protected":false},"author":2,"featured_media":259598,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[37],"tags":[134956,36507,565,134,111,139,69,51741,2357,69763,95454,2359],"class_list":{"0":"post-259597","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-fitness","8":"tag-audio-true","9":"tag-editor-awise","10":"tag-fitness","11":"tag-health","12":"tag-new-zealand","13":"tag-newzealand","14":"tag-nz","15":"tag-parent_category-health","16":"tag-tag-evergreen","17":"tag-tag-longevity","18":"tag-tag-sweat-science","19":"tag-type-article"},"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts\/259597","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/comments?post=259597"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts\/259597\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/media\/259598"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/media?parent=259597"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/categories?post=259597"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/tags?post=259597"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}![\"The](\"https:\/\/www.newsbeep.com\/nz\/wp-content\/uploads\/2026\/01\/mortality-exercise.jpg\"\/)
The more you exercise, the less likely you are to die early. (Photo: Courtesy of BMJ Medicine.)<\/p>\n![\"\"](\"https:\/\/www.newsbeep.com\/nz\/wp-content\/uploads\/2026\/01\/exercise-and-death.jpg\"\/)
The longevity benefits of exercise only seem to plateau if you \u201cadjust\u201d them for factors like BMI, cholesterol, and blood pressure. (Photo: Alex Hutchinson, based on data from BMJ Medicine)<\/p>\n![\"Activities](\"https:\/\/www.newsbeep.com\/nz\/wp-content\/uploads\/2026\/01\/diferent-exercises.jpg\"\/)
Activities like walking, running, bicycling, and swimming all seem to have different effects on longevity. (Photo: Courtesy BMJ Medicine)<\/p>\n![\"No](\"https:\/\/www.newsbeep.com\/nz\/wp-content\/uploads\/2026\/01\/exercise-variety.jpg\"\/)
No matter how much exercise you do, doing a variety of activities seems to improve health. (Photo: Courtesy BMJ Medicine)<\/p>\n