The surprise isn't that exercise is good for you. Everyone knows that. The surprise is how it works, and what it mostly doesn't do.
Exercise is a message, not just a calorie burn
Most people picture exercise as a furnace: you burn fuel, you lose weight, done. A better picture is a signal. Every time your muscles work hard, they send chemical messages to the rest of your body. Researchers call some of these messengers myokines, signalling molecules released by working muscle. In effect, muscle behaves partly like a gland.
Those signals reach your heart, blood vessels, liver, fat, immune system and brain. Here's what changes:
| System | What exercise does |
|---|---|
| Heart and vessels | The heart pumps more blood per beat, so resting heart rate tends to fall. Blood vessels become more flexible. |
| Blood sugar | Working muscle pulls glucose out of the blood through a route that doesn't need insulin. That is why a walk after a meal blunts the sugar spike, even in people with insulin resistance. |
| Muscle and bone | Load tells them to rebuild stronger. Without load they slowly waste away, faster with age. |
| Brain | Blood flow, growth factors and new connections all increase, especially in regions tied to memory. |
| Mood | Exercise reliably lifts mood and helps with depression, as shown below. |
| Sleep | Regular activity tends to deepen sleep and help you fall asleep faster. |
You are not a furnace that exercise stokes. You are a system that exercise retunes.
The brain gets a workout too
In a 2011 trial, 120 older adults were randomly assigned either to a programme of aerobic walking or to stretching and toning. After a year, the walkers' hippocampus, a brain region central to memory, had grown by about 2%. The stretchers' had shrunk by about 1.4%, the usual rate of age-related loss. The researchers described the walkers' gain as reversing one to two years of ageing in that region.
One trial is not proof that exercise prevents dementia, and the link between that growth and memory scores was modest. But it is a striking picture: a structure everyone assumed could only shrink with age grew with walking.
It's a genuine treatment for depression
In 2024 the BMJ published a large analysis of 218 trials with more than 14,000 participants. Walking or jogging, yoga, and strength training all produced moderate reductions in depression compared with control conditions, and more intense exercise tended to help more. The authors concluded that these forms of exercise could be considered core treatments for depression, alongside therapy and antidepressants.
The honest caveat: the researchers rated their confidence in those results as low to very low, because many of the trials were small or hard to blind (you can't secretly give someone a jog). So treat exercise as a strong addition to care, not a replacement for it.
What it mostly doesn't do: melt fat
Here's the part the fitness industry underplays. The anthropologist Herman Pontzer measured the daily energy use of the Hadza, hunter-gatherers in Tanzania who walk many kilometres a day. In 2012 he reported that, adjusted for body size, they burned about the same total calories per day as sedentary Westerners.
His explanation is that the body partly compensates. When you add exercise, it trims energy spent elsewhere. Other scientists dispute how strong that effect is, and exercise clearly does burn some extra energy. But the practical upshot is widely accepted: exercise is powerful for health and weak for weight loss. Diet does most of the work on the scale. Exercise does most of the work on everything else, and it helps keep weight off once it's lost.
That's actually good news. If you've been exercising and the scale hasn't moved, you haven't failed. Your blood pressure, blood sugar, mood and brain are still collecting the rewards.
How much you actually need
The World Health Organization's 2020 guidelines for adults:
- 150–300 minutes a week of moderate activity (brisk walking, cycling, anything that makes talking a bit harder), or 75–150 minutes of vigorous activity (running, hard cycling), or a mix.
- Muscle-strengthening on two or more days a week, working the major muscle groups.
- Sit less. Replacing sitting with any movement helps.
And the most important line in those guidelines: some activity is better than none. The old rule that exercise only "counts" in bouts of ten minutes or more was dropped.
You also need less than you might think to get the biggest gains. A 2025 analysis in The Lancet Public Health found that people walking about 7,000 steps a day had a 47% lower risk of death over the study periods than those walking 2,000. The curve flattened as steps went up: the jump from very little to moderate matters far more than the jump from moderate to heroic. These are observational studies, so some of the gap reflects people who were already healthier, but the shape of the curve is consistent across studies.
For strength training, a 2022 review in the British Journal of Sports Medicine found the biggest associated drop in risk of death, around 10–20%, at just 30–60 minutes a week.
The biggest health return comes from going from nothing to something.
The takeaway
Exercise works less like a calorie furnace and more like a signal that retunes your heart, blood sugar, muscles, bones, brain and mood. It's a real treatment for depression and can even grow a key memory region. It is weak on its own for weight loss, and that's fine, because its main payoffs aren't on the scale. The largest gains come from moving from very little to moderate: roughly 150 minutes of brisk activity and two short strength sessions a week.
Try this: this week, add a 10-minute walk after your largest meal each day, and do one 20-minute session of bodyweight strength work (squats, push-ups against a counter, lunges). That's the first, steepest part of the curve.
Sources & Further Reading
Sources
- Morris, J.N. et al. "Coronary heart-disease and physical activity of work." The Lancet, 1953 (the London bus drivers and conductors).
- Insulin-independent glucose uptake by contracting muscle (GLUT4): standard exercise physiology, reviewed e.g. in Richter & Hargreaves, Physiological Reviews, 2013.
- Erickson, K.I. et al. "Exercise training increases size of hippocampus and improves memory." PNAS, 2011.
- Noetel, M. et al. "Effect of exercise for depression." BMJ, 2024 (218 trials, 14,170 participants).
- Pontzer, H. et al. "Hunter-gatherer energetics and human obesity." PLOS ONE, 2012; and the later debate over constrained energy expenditure.
- WHO guidelines on physical activity and sedentary behaviour, 2020 (Bull et al., British Journal of Sports Medicine).
- Ding, D. et al. "Daily steps and health outcomes in adults." The Lancet Public Health, 2025.
- Momma, H. et al. Meta-analysis of muscle-strengthening activities and risk of death and disease. British Journal of Sports Medicine, 2022.
Go deeper
- 📘 Burn — Herman Pontzer (the Hadza, metabolism, and why exercise isn't about calories)
- 📘 Exercised — Daniel Lieberman (why humans evolved to move but not to "exercise")
- 📘 Spark — John Ratey (exercise and the brain, for a general reader)
- 🌐 WHO — "Every move counts" guidelines announcement
Sensitive-topic note: this is general information, not medical advice. If you have a heart condition, chest pain on exertion, are pregnant, or are starting after a long break with health concerns, check with a doctor first. For depression, exercise works best alongside, not instead of, professional care.