Menopause Bone Loss Has a Metabolic Clue: Insulin Resistance

Estrogen loss helps explain menopause-related bone loss, but it does not explain the whole picture. In a large longitudinal study of midlife women, a faster rise in insulin resistance tracked faster lumbar-spine bone loss before, during, and after the menopause transition. The association was largest during the transition itself. That does not prove insulin resistance causes bone loss. It does suggest that bone health and metabolic health may be moving in the wrong direction together.

The practical implication is less exotic than the biology: long-term combined exercise, built around resistance work and supplemented with impact exercise when appropriate, has the best evidence for preserving bone mineral density. It may also improve glucose regulation. One training plan can plausibly address both risks, even though no trial has yet proved that improving insulin resistance slows menopausal bone loss.

Insulin resistance tracked bone loss most closely when it worsened

Shieh et al.'s 2022 SWAN analysis followed multiethnic midlife women across menopause stages: 861 premenopausal women, 571 in the menopause transition, and 693 postmenopausal women. The investigators used repeated measures and piecewise mixed-effects models to examine HOMA-IR, a fasting glucose-and-insulin estimate of insulin resistance, alongside changes in lumbar-spine and femoral-neck bone mineral density (BMD).

The relationship was nonlinear. The researchers identified an inflection point at a HOMA-IR of about 2.82. Below it, higher HOMA-IR was associated with slower BMD decline at selected sites. For example, each doubling of HOMA-IR was associated with 0.0032 g/cm² per year slower lumbar-spine loss in premenopause (P=0.01) and 0.0042 g/cm² per year slower lumbar-spine loss after menopause (P=0.004). This should not be read as insulin resistance protecting bone. Greater body weight can increase mechanical loading on the skeleton, and that likely explains at least part of the apparent advantage.

Above HOMA-IR 2.82, the absolute value no longer predicted BMD change. The rate of worsening did. Each standard-deviation faster annual increase in log₂ HOMA-IR was associated with an additional 0.0026 g/cm² per year of lumbar-spine loss in premenopause (P=0.001), 0.0081 g/cm² per year during the menopause transition (P=0.005), and 0.0021 g/cm² per year after menopause (P=0.03). Between 19.7% and 29.4% of participants exceeded the threshold at a given stage, with prevalence rising through the transition.

The study adjusted for age, race and ethnicity, smoking, body weight, and weight change. It did not adjust for estradiol, visceral fat distribution, or lean mass. Those omissions matter because all three relate to both insulin resistance and bone. SWAN found an association, not a metabolic smoking gun, and its HOMA-IR threshold has not been externally validated. Still, the pattern makes a reasonable case for watching the trajectory of metabolic health during menopause rather than treating bone as an estrogen-only problem.

The best bone program is combined and long enough to matter

A 2026 network meta-analysis pooled 74 randomized controlled trials involving 5,331 postmenopausal women. Every intervention lasted at least 24 weeks. Combined mind-body and resistance exercise ranked first for BMD at the lumbar spine, femoral neck, Ward's triangle, and total body. Multimodal exercise also produced consistent benefits at the total hip and femoral neck. By contrast, resistance, impact, aerobic, and mind-body exercise used alone produced effects limited to particular skeletal sites; none produced a significant intertrochanteric effect.

Duration changed the interpretation. Subgroup and meta-regression analyses found that programs lasting at least 52 weeks produced more consistent, sustained BMD improvements. Effect sizes declined over time at most sites, so a short trial may not tell us much about what a year of training does. Bone remodeling works on a slower clock than a six-week fitness challenge.

An umbrella review published in 2024 reached a similar conclusion: combined resistance protocols and concurrent training preserved BMD at the femoral neck, lumbar spine, Ward's triangle, and trochanter in menopausal women. Aerobic exercise alone and resistance exercise alone produced inconsistent BMD results. The evidence supports a combined program more strongly than a claim that lifting by itself reliably solves menopause-related bone loss.

Impact work adds information that a DXA scan can miss

BMD is useful, but it is not the entire engineering report on a bone. A 2023 systematic review and meta-analysis of randomized trials found that moderate- to high-impact training improved structural measures including cortical thickness, cross-sectional area, and section modulus across the lifespan, independently of areal BMD. Menopause-specific structural data in that review were limited, so this is supportive rather than definitive evidence for menopausal women.

For someone medically cleared for it, impact exercise can complement resistance-based training rather than compete with it. The appropriate version depends on fracture risk, joint symptoms, prior injuries, and current fitness. Jumping into plyometrics with untreated osteoporosis is less a training plan than an argument with gravity.

The metabolic case is promising, but indirect

The exercise prescription has a second appeal: resistance and combined exercise rank highly for glycemic control and HOMA-IR reduction in people with prediabetes. In a 2022 network meta-analysis, those modalities ranked highest among exercise approaches for these metabolic outcomes. But the participants were not stratified by menopausal stage, so the result cannot establish the same effect size in perimenopausal or postmenopausal women.

A 2024 systematic review and meta-analysis also found that resistance training improves glucose homeostasis. Skeletal muscle adaptations, including changes in muscle mass and glucose transport capacity, offer a plausible explanation. Yet gains in fat-free mass did not consistently correlate with reductions in HOMA-IR across the included trials. More muscle may help, but the evidence does not support treating hypertrophy as a guaranteed route to better insulin sensitivity.

Aerobic exercise also deserves a place in the conversation. A 2024 meta-analysis of randomized trials in postmenopausal women found improvements in cardiometabolic risk factors, including body composition, lipids, and blood pressure. In another randomized trial, Ruegsegger et al. found that high-intensity aerobic training, but not resistance or combined training, improved both cardiometabolic health and skeletal-muscle mitochondrial dynamics. That study was not menopause-specific, but it is a useful warning against declaring one modality the universal metabolic winner.

Why one prescription will not fit every insulin-resistant woman

Insulin resistance is not one uniform condition. A 2025 systematic review and meta-analysis found that lifestyle responses differed between impaired-fasting-glucose and impaired-glucose-tolerance phenotypes. The first more often reflects hepatic insulin resistance; the second more often reflects peripheral, skeletal-muscle-based insulin resistance. Menopausal exercise trials have not sorted participants this way. A resistance-centered program may therefore help some metabolic profiles more than others.

That limitation does not weaken the structural recommendation. Combined resistance-based exercise has the best current evidence for preserving BMD at key sites, particularly when maintained for at least a year. It does limit the larger claim: nobody has run the trial that randomly assigns menopausal women to exercise, documents a reduction in insulin resistance, and then shows that reduction mediates less bone loss.

A sensible evidence-based plan

For menopausal women who can exercise safely, prioritize a sustained combined program centered on resistance training, add suitable impact loading, and keep aerobic work in the mix for cardiovascular and metabolic health. Plan in years, not weeks. DXA remains the clinical tool for assessing BMD and fracture risk in the appropriate setting.

Monitoring HOMA-IR over time may offer useful metabolic context, particularly when it is rising, but HOMA-IR 2.82 is not a validated treatment threshold. The strongest conclusion is narrower and more useful: combined, long-term training is well supported for bone preservation, and it may efficiently address a metabolic risk factor that tracks with bone loss. The causal bridge remains an open research question.