Melatonin Plus Bone Loading: A Promising but Early Trial

A small spine signal, not a bone-building verdict

A 5 mg nightly melatonin supplement may have helped an osteogenic-loading program improve lumbar-spine measurements in postmenopausal women with osteopenia. That is an interesting result. It does not show that melatonin prevents fractures, improves the hip, or reliably boosts the effects of exercise.

The direct evidence comes from the 12-month MelaOstrong pilot randomized controlled trial. In the osteogenic-loading-plus-melatonin arm, lumbar-spine T-score increased by a mean 0.13 (SD 0.05; p=0.015) and lumbar-spine bone mineral density (BMD) increased by 0.013 g/cm2 (SD 0.006; p=0.021) from baseline. The accessible report did not show significant between-group T-score differences, nor significant improvement at the hip or femoral neck, or in FRAX estimates of fracture risk.

That distinction controls the conclusion. A statistically significant change within one treatment arm can generate a worthwhile hypothesis. It cannot, on its own, establish that melatonin amplified loading more than loading alone, placebo, or a mock-loading control.

The current evidence supports an experimental idea: build the program around appropriately prescribed mechanical loading, and regard melatonin as an unconfirmed add-on.

What the trial actually tested

MelaOstrong studied postmenopausal women with osteopenia and combined nightly melatonin with a loading protocol designed to apply progressively greater mechanical force than a mock-loading condition. The program targeted the upper body, core and spinal stability, lower body and hip, and posture-related regions. The study reported forces up to 4.2 times body weight.

The outcome that moved was lumbar-spine BMD and T-score within the combination arm over 12 months. The publication record available for this trial does not provide usable details needed to judge the result fully, including sample size, allocation and blinding procedures, adherence, attrition, adverse-event counts, session frequency, or DXA precision error and least-significant-change thresholds.

That last omission matters. A 0.013 g/cm2 gain points in the favorable direction, but DXA measurements have unavoidable precision limits. Without the scanner's precision data, readers cannot tell whether the average change clearly exceeded measurement noise. The result is therefore statistically notable within the study, while its clinical size remains uncertain.

Why loading remains the more established part of the stack

Exercise has a credible physiological case in postmenopausal bone health. A 2025 systematic review and meta-analysis of 24 randomized trials, including 1,067 postmenopausal women, found that exercise increased alkaline phosphatase, P1NP, and osteocalcin while reducing CTX. Those biomarkers fit a shift toward bone formation and away from resorption.

Biomarkers do not equal fewer fractures, and they do not guarantee a large DXA response. Still, they help explain why loading belongs at the center of the idea rather than being treated as scenery for a supplement.

Some higher-impact activities have produced measurable, site-specific results. In a 16-week randomized trial of inactive postmenopausal women, recreational team handball increased lumbar-spine BMD by about 1.5% and lumbar-spine bone mineral content by about 2.3%, alongside large increases in formation markers. A systematic review of recreational football found a lower-limb BMD mean difference of 0.02 g/cm2 (95% CI, 0.00 to 0.03; p=0.02), while total-BMD effects were negligible.

These findings make a lumbar-spine-specific response to an osteogenic program plausible. They also show why broad claims about "building bone everywhere" outrun the evidence. Bones respond to the loads they experience; trials often find effects at particular skeletal sites, not across the whole skeleton.

Why the melatonin claim remains unproven

Melatonin has one direct human pilot signal in this specific combination. That is not the same as a replicated treatment effect. MelaOstrong did not report a significant between-group T-score difference, which is the comparison most relevant to the claim that melatonin enhanced loading. It also did not report significant hip, femoral-neck, or FRAX improvement.

The exploratory correlations in the combination arm add a possible clue but not a mechanism. Timed Up and Go performance correlated strongly and negatively with CTX; sit-to-stand change correlated negatively with the P1NP:CTX ratio and positively with melatonin. Correlations with absolute values around 0.96 to 0.98 can look dramatic in a pilot dataset, but they can also shift sharply with a few observations. They do not prove that melatonin changed bone remodeling or performance through a particular pathway. CRP and cortisol did not change in the accessible record, so the trial does not support a clear anti-inflammatory explanation.

The broader literature offers a useful warning against mechanism-first enthusiasm. A 2024 meta-analysis of probiotic trials concluded that proposed bone benefits still needed validation. Then a 48-week, multicenter, double-blind randomized trial in 72 postmenopausal women found that a probiotic combination produced no change in bone-health outcomes. A sensible biological story is an invitation to test an intervention, not a substitute for a successful test.

Results also vary with the intervention and endpoint. A 12-week aerobic-exercise trial in 25 women per group with postmenopausal osteopenia found no significant between-group changes in P1NP, CTX, osteocalcin, or alkaline phosphatase, despite some within-arm improvements. A different loading prescription, duration, comparator, or measured site can reasonably produce a different result; those differences should not be blended into a single oversized promise.

Surrogates are useful, but fractures decide the clinical question

BMD, T-scores, P1NP, osteocalcin, alkaline phosphatase, and CTX are useful measures of bone status or remodeling. None is a fracture outcome. MelaOstrong measured no fractures and found no FRAX improvement, so it cannot establish fracture-risk reduction.

A 2025 network meta-analysis of dietary supplements and bone-turnover markers, covering 43 randomized trials and 5,184 postmenopausal women, illustrates the same problem: marker rankings can identify candidates worth studying, but they do not answer whether a supplement prevents fractures. The abstract highlighted vitamin E, vitamin D plus K, vitamin K, calcium, and protein among the top-ranked biomarker interventions—not melatonin.

Safety also needs better reporting before this becomes a casual do-it-yourself protocol. The primary trial's accessible record does not provide a usable adverse-event table or resolve practical questions about next-day sedation, medication interactions, hormone-sensitive conditions, or injury risk from high-force loading. Absence of a reported problem is not evidence that the problem was assessed and ruled out.

The practical verdict

Postmenopausal women with osteopenia should not read one lumbar-spine finding as a reason to replace clinical assessment, osteoporosis treatment when indicated, or a well-designed exercise plan with a bottle of melatonin. The trial tested 5 mg nightly melatonin alongside loading; it did not establish melatonin as a standalone bone treatment or a fracture-prevention strategy.

The reasonable interpretation is narrower. Osteogenic loading has support as a bone-health strategy, although its effects are often modest and site-specific. Melatonin has a biologically plausible role and one promising 12-month pilot result in combination with loading. A larger preregistered trial with transparent allocation, adherence and safety reporting, prespecified between-group analyses, DXA precision thresholds, and fracture-relevant follow-up would tell us whether that signal holds up.

Until then, the stack deserves curiosity, not certainty.