We know exercise is healthy — but proving it causes better health is surprisingly hard
Everyone knows that fit people live longer and get sick less often. But proving that exercise actually causes that better health — rather than the other way around — turns out to…
The problem is a classic one in epidemiology: people who exercise regularly also tend to eat better, have less stress, higher incomes, and better access to healthcare. How do you isolate exercise’s contribution from all those other factors? The method increasingly used is called Mendelian randomization. It works by using genetic variants naturally associated with a trait — in this case, physical fitness — as a kind of ‘natural experiment’. Because your genes are fixed at birth and are not influenced by your lifestyle choices, genetic variants can serve as a neutral proxy for the trait.
What genes reveal about fitness and health
The researchers identified genetic variants that predict how naturally fit a person tends to be — then looked at whether people carrying those variants are also healthier. The logic: if fitness genuinely causes better health, people who are genetically predisposed toward higher fitness should have better health outcomes regardless of their actual behavior. The findings do support a causal relationship between physiological fitness and lower rates of disease and mortality.
But the method has limits. Genetic variants for fitness encompass traits like muscle mass, lung capacity, and cardiovascular efficiency — not just the tendency to exercise. What the study actually measures, in part, is innate biological capacity, not the habit of regular physical activity per se. Someone with genetically high aerobic capacity who never exercises, and someone with low natural capacity who trains hard, are not simply comparable within this model.
Why this still matters
That exercise is good for you was never really in doubt. The value of this type of research lies not in the conclusion itself, but in the strength of evidence behind it. For policy and clinical guidelines, there is a meaningful difference between demonstrating a correlation and demonstrating causation. Mendelian randomization brings us closer to that causal proof than purely observational research — but it doesn’t replace the gold standard of a randomized controlled trial. And that, for long-term health benefits of exercise in humans, is nearly impossible to conduct.