Multiple randomised studies and meta-analyses point consistently in the same direction: high-intensity resistance training improves bone density in postmenopausal women more effectively than low-intensity programmes. The evidence is based on multiple randomised trials and meta-analyses, but the studies are predominantly small to medium in size and the between-study variation is large. In practice, this means that a well-supervised heavy resistance programme (above 70-85% of one-repetition maximum) is the best choice for bone preservation, and that a combined programme is the appropriate option when fall risk and mobility concerns are also present.
High-intensity resistance training currently has the strongest and most concrete evidence base when it comes to preserving bone in later life. The LIFTMOR study, a randomised controlled trial in 101 postmenopausal women with osteopenia or osteoporosis (mean age 65 years), showed that training at more than 85 percent of one-repetition maximum, twice a week for eight months, improved lumbar spine bone density by almost 3 percent, while it declined by 1.2 percent in the control group. Femoral neck bone density also improved significantly. Adverse effects were minimal: one participant experienced a minor back spasm. This study is relatively small and lasted eight months, so the findings should be interpreted with caution, but the direction is clear1.
A meta-analysis of seventeen randomised studies (690 participants) confirms that resistance training improves bone density at the lumbar spine, femoral neck and total hip in postmenopausal women. The subgroup analysis points to three key factors: an intensity of at least 70 percent of one-repetition maximum, three sessions per week and a programme lasting at least 48 weeks. Effect sizes were medium to large. One caveat: the variation between the individual studies was high (I²=87-91%), which means the pooled outcome should be read with some caution2.
For older women with osteoporosis who want to address several problems at once, a combined programme of strength, aerobics, balance and flexibility is also well supported by evidence. A meta-analysis of fourteen randomised studies (544 participants, mean age 68.4 years) showed improvements in bone density, muscle strength, balance and quality of life, as well as a lower risk of falls. Exact figures were not reported in the abstract. The advantage of this type of programme is that it addresses multiple risk factors for fractures simultaneously, not only bone density3.
In very old and/or obese women (mean age 80 years), the primary goal appears to shift from improving bone to preventing further decline. In a small study (30 participants) with 24 weeks of resistance training twice a week, bone density did not improve significantly in the training group, but it did decline significantly in the control group. In addition, muscle strength and functional fitness clearly improved. This suggests that preventing bone loss at advanced age is already a meaningful outcome, even when absolute gains in bone do not occur4.
Creatine supplementation most likely adds nothing to resistance training in terms of bone. A review article concludes that most studies in older adults show no significant improvement in bone mass with creatine. There are some indications of muscle strength gains in combination with training, but for bone specifically the evidence is not convincing5. Clinical guidelines emphasise that exercise is part of a broader package, alongside adequate calcium, vitamin D and, where indicated, medication; the direct evidence that training on its own prevents fractures is still limited and partly based on expert opinion6,7.
Two meta-analyses (PMID 36361073, 40420105), one RCT (PMID 28975661), one small controlled trial (PMID 36361434), one narrative review (PMID 35055695), one review on creatine (PMID 39509039) and two clinical guidelines (PMID 35478046, 40587168). Total participants estimated on the basis of the reported n values: 101 (LIFTMOR) + 690 (resistance meta-analysis) + 544 (multicomponent meta-analysis) + 30 (obese older adults) = ~1365; narrative reviews and guidelines are not counted.