The cyclist's bone problem: why riding alone doesn't build a skeleton

Cycling is very good for almost everything. It is one of the few sports you can do at high volume for decades without wrecking your joints, it does more for cardiovascular health than nearly anything else available, and it is gentle enough that people ride well into their eighties. The skeleton is the exception, and it is a large enough exception to be worth a whole article.

Road cyclists, as a group, have lower bone mineral density than people who do not train at all. Not lower than runners — lower than sedentary controls of the same age. That finding turns up repeatedly in systematic reviews, and it is uncomfortable, because it means the training itself is part of the problem rather than merely failing to help.

This is not a reason to stop riding. It is a reason to add about forty minutes a week of something else, and to stop treating the off-season as a slightly shorter version of the season.

What the data actually shows

Systematic reviews of bone health in cyclists converge on the same picture. Adult road cyclists — particularly those training at high volume, and particularly professionals — show lower bone mineral density than non-athletic controls, with the deficit most pronounced at the lumbar spine. A longitudinal study of competitive male cyclists found bone mineral density decreasing over the course of a single training year. A seven-year follow-up of male master cyclists found declines that their non-athletic counterparts did not show.

Two honest caveats. Most of this work is cross-sectional, so it cannot fully separate "cycling lowers bone density" from "people with a particular build are drawn to cycling". And the strongest effects come from professional and high-volume populations, not from someone riding eight hours a week. The direction is consistent across reviews; the size of the effect for a given amateur is not something anyone can quote you.

Why riding does not build a skeleton

The first of those is the structural one. Bone remodels in response to mechanical strain, and the signal it listens for is strain that is high in magnitude, applied at a high rate, and unfamiliar. Position stands on physical activity and bone health describe exactly that: impact activities and resistance training at relatively high loads, not endurance activity at high volume. Cycling is the near-perfect example of the loading pattern bone ignores — your body weight is on the saddle, the force curve is smooth, and the movement is identical several hundred thousand times a year.

The second is less widely known and more specific to long rides. You lose calcium in sweat, and on a hot four-hour ride that loss is enough to reduce circulating ionised calcium. The body defends blood calcium above almost everything else, so parathyroid hormone rises and calcium comes out of the skeleton to cover the shortfall. Controlled work has shown that taking calcium before exercise attenuates both the parathyroid hormone rise and the rise in bone resorption markers — which tells you the mechanism is real, even if the long-term consequence has not been demonstrated directly.

The third multiplies the other two. Low energy availability suppresses the endocrine environment bone depends on, and low bone mineral density is one of the defined outcomes in the international consensus on relative energy deficiency in sport. If you are riding a lot, sweating a lot and eating less than you need, all three mechanisms are running at once. Our articles on low energy availability and REDs and under-fuelling cover that side in detail, and it is the part with the shortest route to harm.

Why it matters: the crash

A sedentary person with slightly low bone density spends their life not falling over at 50 km/h. A cyclist does not have that luxury. The combination that makes this worth acting on is not the density number by itself — it is low bone density in a sport where hitting the ground hard is a routine occupational hazard.

Here the evidence thins out, and it is worth saying so. There is good data that cyclists have lower bone density than controls. There is much less data directly linking that to fracture rates in cyclists specifically, because the studies would be difficult and nobody has done them well. The link between low bone density and fracture risk in the general population is firmly established, and it is reasonable to expect it to carry over. But if someone tells you a specific percentage by which cycling raises your fracture risk, they are extrapolating.

What actually helps

If you only do one of those, do the lifting. It has the best evidence, it fits into a cycling week without much disruption, and the case for it stands on its own even if you had no skeleton to worry about — why every cyclist should lift weights sets out the performance half of the argument. If you can do two, add impact: a few sets of hops and skips before the lifting session takes five minutes and covers the loading pattern that squats alone do not.

What the dose looks like in a real week

Bone is a slow tissue with a saturating response, and that combination is good news for a busy rider. The loading signal saturates after a fairly small number of repetitions, which means more is not better past a certain point — five sets of ten hops does very little that two sets of ten does not. What matters is that the loading is heavy, quick and repeated across the year rather than crammed into January.

In practice that comes to about forty minutes a week: two gym sessions of twenty to thirty minutes, each opening with two or three sets of hops or skips before the heavy lifts. Put them on days that are already hard so they do not eat into recovery days. That is the entire intervention, and it is small enough that the usual objection — no time — does not really survive contact with the arithmetic.

Bone also responds on a slower clock than muscle. Expect to see nothing for a year, which is precisely why it gets abandoned. Judge it by whether you are still doing it in three years, not by how you feel in March.

Who should get a scan

A DXA scan is not something every cyclist needs. It becomes worth discussing with a doctor if you tick more than one of: a history of stress fracture or a fracture from a low-energy fall; years of high-volume riding with no other sport; a period of deliberate weight loss or disordered eating; loss or disruption of menstrual periods; or being a masters rider who has only ever cycled.

The number that comes back is a Z-score, comparing you with others of your age and sex, and the threshold generally used for athletes is −1.0 rather than the −2.5 used for osteoporosis in the general population. Athletes are expected to have better bone than average, so merely average is already a finding.

One last thing about timing. Bone responds slowly — you are looking at changes over years, not blocks. That means the cyclist with the most to gain from starting is the one in their twenties and thirties who feels entirely fine, and the one most likely to act is the one in their sixties who has just broken a hip. The changes that arrive after 40, 50 and 60 are easier to live with when the skeleton underneath them was looked after early.