Exercise and Bone Density — What Does a New Network Meta-Analysis Show?

Polish version

Key message: regular exercise can help limit age-related loss of bone mineral density, but different forms of activity are not equally effective and the evidence does not support one universal programme for everyone. A new network meta-analysis suggests the strongest signals for brisk walking or jogging and for programmes combining aerobic and resistance exercise.

Why is this study interesting?

Advice to “be more active” is broadly correct but not very specific. In a paper published in BMJ on 9 September 2026, the authors asked two practical questions: which forms of exercise best preserve bone mineral density after the age of 40, and what weekly exercise dose may be needed to produce a measurable effect?

Question and methodology

The investigators performed a hierarchical Bayesian network meta-analysis together with a dose-response analysis. The review included 124 randomised trials involving 18,429 participants aged at least 40 years. Structured exercise programmes were compared with no exercise or with other forms of activity.

The main outcomes were changes in bone mineral density at the lumbar spine, femoral neck and total hip. Fractures were assessed as an additional outcome. The authors also modelled the relationship between weekly exercise dose, expressed as MET-minutes, and bone-density effects.

Main results

• At the lumbar spine, brisk walking or jogging and programmes combining aerobic with resistance exercise were associated with an average improvement in bone mineral density of about 0.013 g/cm² compared with control.
• At the femoral neck, favourable signals were seen for brisk walking or jogging and for exercise combining movement, balance and concentration, such as tai chi.
• For total hip bone mineral density, brisk walking or jogging ranked highest, but certainty in that estimate was low.
• The dose-response relationship was not linear. The threshold for a potentially clinically meaningful benefit was estimated at about 400 MET-minutes per week for the lumbar spine and about 600 MET-minutes for the femoral neck and total hip.
• The fracture analysis included 26 trials and 11,132 participants. Mixed aerobic exercise and mind-body exercise were associated with lower fracture risk, but certainty in these data was low.

Clinical interpretation for an orthopaedic surgeon

The most useful conclusion is not that everyone should start running. The findings support programmes that include skeletal loading, aerobic activity and strength training, adapted to age, physical capacity, fall risk and fracture risk. Around 600 MET-minutes per week corresponds roughly to 2–3 hours of brisk walking or jogging, but this is an average signal derived from a statistical model, not a ready-made prescription for an individual patient.

Jogging may be inappropriate for someone with advanced osteoporosis, a recent fracture, impaired balance or substantial joint disease. In such situations, safe resistance training, balance work and graded loading may matter more than mechanically trying to achieve one numerical exercise target.

Limitations

• The included trials differed in population, duration, exercise intensity and monitoring of adherence.
• A substantial proportion of comparisons between exercise methods were indirect, as is typical in network meta-analysis.
• Certainty for many estimates was low or moderate, and rankings of interventions may be less stable than individual numbers suggest.
• Bone mineral density is an important surrogate outcome but is not equivalent to a demonstrated reduction in fractures.
• Fracture data were less abundant than densitometry data and therefore provide less certain evidence about actual fracture prevention.

What could this change in practice?

The study strengthens the case for making exercise advice for prevention of bone loss more specific. Recommendations should address the type of exercise, its weekly dose and safety. A sensible programme will often combine weight-bearing lower-limb activity with resistance training and balance work.

What does the study not prove?

The meta-analysis does not prove that walking or jogging alone prevents fractures in every patient. It also does not show that exercise replaces osteoporosis assessment or pharmacological treatment in people at high fracture risk. Nor does it justify naming a single form of activity as best regardless of age, comorbidity and functional capacity.

Back to: For Doctors and Physiotherapists

Source

Lu C, Li L, Zhang L, et al. Effect of exercise on bone health in middle aged and older adults: hierarchical network meta-analysis of randomised trials. BMJ. 2026;394:e100561. DOI: 10.1136/bmj-2026-100561. PubMed.

Educational material for healthcare professionals. It does not constitute an individual therapeutic recommendation.

Content prepared by: Tomasz Poboży - orthopaedic surgeon.