Bone needs impact and fuel, not just miles
High confidence
Bone gets stronger from loading that is punchy, varied, and gradually harder. Hills, faster running, jumps, and strength work do more for it than piling on easy miles. And none of that works if you're under-fueling. Too little energy stalls bone repair and is a leading reason runners get stress fractures. Building bone is about the kind of load and eating enough, not just the mileage.
Why it works
Bone responds to loads that hit hard, hit fast, and feel unfamiliar. Repetitive low-impact running sends only a weak signal to build new bone. Rebuilding bone, called remodeling, also needs enough energy, calcium, vitamin D, and normal hormones. When energy availability runs low, meaning you eat too little to cover your training, bone breaks down faster than it rebuilds and stress-injury risk climbs.
What it means in practice
This is why base building is not only about volume. Progressive mileage builds bone best when it comes with some higher-impact or resistance work and enough fuel. Low-energy-availability risk is highest in runners logging heavy volume or ramping up fast. Pair with stress-fractures-need-multifactorial-management and adaptation-bone-slowest-months.
The evidence
-
Both ground-reaction impact and muscle contraction stimulate bone. The stimulus that works is dynamic rather than static, high in magnitude and rate, and specific to the loaded site. Cross-sectional data show the highest BMD in high-impact athletes such as dancers, gymnasts, and soccer players, while swimmers and cyclists sit at or below controls. Intensity and novelty matter more than duration, and a few loading cycles appear to be enough; the evidence is cross-sectional, so it shows association rather than proof that the loading built the bone.
-
The guideline concludes that the intensity and novelty of a load drive bone strengthening more than any other factor. Bone responds most to short periods of unusual strain distribution, high magnitude, and a rapid loading rate. In a 12-month multimodal program, high-intensity resistance training paired with a weight-bearing circuit of moderate-impact activities produced significant BMD improvement at the femoral trochanter.
-
A meta-analysis of nine controlled trials in premenopausal women found impact loading significantly raised lumbar spine and femoral neck aBMD. Impact-only work reached the femoral neck; adding resistance to impact reached several sites. A 2019 study followed 90 female army recruits through eight weeks of basic combat training and found trabecular bone density rose through thicker trabeculae, with cortical gains concentrated in the most-loaded regions.
-
Low energy availability hurts bone. This review finds it lowers bone-formation markers (P1NP) and raises resorption markers (beta-CTX), and it tracks with lower BMD and higher stress-fracture risk. As background, the authors note bone responds most to high-impact multidirectional loading at weight-bearing sites, while low-impact repetitive endurance running or non-weight-bearing sport does little for it. This is a narrative review, and the loading hierarchy is background framing rather than its own contribution.
-
The runners carried higher BMD Z-scores than controls at the proximal femur (1.30 vs 0.20) and total body (1.70 vs 0.90), but the lumbar spine showed no difference. Forty-seven percent of them screened at risk of low energy availability, and proximal-femur BMD rose with estradiol and fell with low-energy-availability symptoms. So the bone benefit follows where running loads the skeleton, and low energy availability eats into it. This is one small cross-sectional study of elite women, so it maps associations rather than proving cause.
n=30
-
In endurance-trained individuals, animal-protein intake correlated with tibial cortical bone strength and density, while higher total protein tracked with higher lumbar-spine aBMD. As background, the paper frames endurance runners as prone to impaired bone health and elevated bone-stress-injury risk from low-magnitude repetitive loading, high training volume, and low energy availability. It also notes their whole-body and site-specific aBMD runs lower than in higher-impact-sport athletes. The design is cross-sectional and leans on diet recall, so it shows association, not that more protein builds bone.
-
The IOC panel ties low energy availability to impairment across multiple systems, bone health among them, alongside menstrual function, metabolic rate, immunity, protein synthesis, and cardiovascular health. It names low energy availability as the upstream driver that links under-fueling to weaker bone and higher stress-injury risk in athletes. This is expert consensus rather than primary data.
-
A prolonged continuous treadmill run produced no true bone-formation response. The transient P1NP rise was a biological artefact, likely connective-tissue leakage rather than new bone. Beta-CTX-1 changes tracked circadian rhythm, not the run itself. This reinforces that repetitive endurance running is a weak direct osteogenic stimulus, consistent with its links to stress-fracture risk and low BMD at non-loaded sites.
-
Walking has only a small effect on spine bone density, but a significant positive effect on the femoral neck in postmenopausal women. Maintaining spinal bone takes targeted weight-bearing and high-intensity resistance loading, not repetitive low-impact activity. In women aged 75-85, cortical volumetric BMD at the radius rose after 6 months of twice-weekly resistance training (Liu-Ambrose), with training duration the strongest predictor of bone parameters. As a narrative review, it maps the direction of effect rather than pinning exact doses.
-
Over 12 months, both the low- and high-strain-rate groups gained ultradistal bone mineral content (+2.7% and +3.4%), while controls lost bone (-1.3%). The largest gains showed up in women completing about 128 loading bouts at roughly 575 microstrain. Loading dose (magnitude, rate, and number of bouts) tracked positively with bone change. But it explained under ~15% of the variance, so individual response varied widely.
Why we call confidence high
Reviews and trials agree that bone-building loads (osteogenic loads) must be dynamic, high in magnitude and rate, and novel rather than static or purely repetitive (Morseth, Izquierdo, Mancuso, Papageorgiou). Impact and higher-intensity work produce larger bone responses than low-impact endurance activity. Swimmers and cyclists often show lower bone mineral density (BMD) than controls (Morseth). Endurance runners show site-specific bone benefit but stay vulnerable at less-loaded sites and under low energy availability (Kyte, Gardy, Civil). Low energy availability is a keystone modifier (Mountjoy RED-S consensus, Papageorgiou), and acute running alone is a weak direct signal to build bone (Civil). So miles without impact variety and fueling build bone poorly.
Where it applies
Adult runners. Low-energy-availability risk is most acute in female runners (Female Athlete Triad / RED-S), but male endurance athletes are affected too.
Does not apply to: clinical osteoporosis pharmacotherapy.
Last reviewed Jun 29, 2026. See how we score.