Which supplements help bone density in perimenopause?

In the UK, meet the intakes for calcium (700 mg a day) and vitamin D (10 micrograms) first, from food where you can. Those two carry authorised bone claims and the largest trials. Vitamin K2 and creatine on its own did not change bone density in trials of two to three years, and the direct evidence in perimenopause itself is thin.

• Calcium is needed for the maintenance of normal bones and vitamin D contributes to it. Both are authorised claims. The UK reference nutrient intake for calcium is 700 mg a day; the figure on labels is 800 mg.
• Vitamin K2 at 375 micrograms a day for three years, on top of calcium and vitamin D, shifted a bone marker but did not change bone density or bone structure compared with placebo.
• The trials here are in postmenopausal women. For perimenopause the same nutrients apply, and loading exercise and a fracture-risk assessment matter alongside any supplement.

Why this question comes up in this window

Bone loss speeds up in the years either side of the final period. That is why the question of what helps bone density stops being abstract in perimenopause, and it is the reason this page exists.

One thing to say first: nearly all of the trial evidence below is in postmenopausal women, some of them selected for already-low bone density. The nutrients are the same in perimenopause, but the direct evidence for this stage is thinner, and the page says so where it applies.

Calcium and vitamin D

Calcium is needed for the maintenance of normal bones. Vitamin D contributes to the maintenance of normal bones and to the normal absorption and use of calcium. Both are GB-authorised claims.

The UK reference nutrient intake for calcium in adults is 700 mg a day [7]. The figure printed on supplement labels is the 800 mg nutrient reference value used for labelling across Europe, which is why the two numbers differ. Seven hundred milligrams is reachable from food for most people: roughly a pint of milk, or equivalent amounts of yoghurt, cheese, tinned fish with the bones, calcium-fortified plant milks or leafy greens [7]. The UK vitamin D advice is 10 micrograms (400 IU) a day, and our vitamin D page covers who should take it year-round.

On fractures the picture is contested. The largest trial gave 36,282 postmenopausal women 1,000 mg of calcium with 400 IU of vitamin D3 daily for an average of seven years. Hip bone density was 1.06 per cent higher. Hip fractures were not significantly reduced: the hazard ratio was 0.88, with a confidence interval from 0.72 to 1.08. A hazard ratio below 1 means fewer events, and an interval that includes 1 means the difference could be chance. Kidney stone risk rose. A secondary analysis limited to the women who kept taking the tablets gave a hazard ratio of 0.71, which is not the same as the randomised comparison [1]. A meta-analysis across eight trials and 30,970 people reported a 15 per cent relative reduction in total fractures and 30 per cent in hip fractures [2], while noting that it draws on a subgroup of that same trial.

A reasonable summary: meet the reference intakes, preferably through food, with a supplement filling any gap. Not a fracture-proofing strategy on its own.

Tested, and did not change bone density

Vitamin K2. A three-year randomised trial gave 142 postmenopausal women with osteopenia either 375 micrograms of MK-7 daily or placebo, both groups also receiving calcium and vitamin D. Carboxylation of osteocalcin rose in the K2 group, which is the mechanism proposed for it. Bone density fell at every site measured, with no difference between groups, and bone structure and turnover markers did not differ either [3]. The marker moved; bone density did not.

Creatine on its own. Two years at 3 g a day in 200 postmenopausal women with osteopenia: no difference from placebo on bone density, bone markers, lean mass, physical function, falls or fractures [4].

Both are properly run trials with null results on bone density.

Where there is a signal, with the caveats attached

Creatine added to resistance training. Forty-seven postmenopausal women were randomised to twelve months of supervised training three days a week with either creatine at 0.1 g per kg of body weight daily or placebo. Femoral neck bone density fell 1.2 per cent with creatine against 3.9 per cent with placebo, and the width of the femoral shaft increased [5]. Both groups trained, so the difference is what creatine added to the training. Thirty-three women completed the study, and the dose, roughly 7 g a day for a 70 kg woman, is higher than the 3 g used in the authorised claim for creatine. Our creatine page covers this in more detail.

Collagen peptides. Five grams daily for twelve months in 131 postmenopausal women raised spine and femoral neck T-scores against control [6]. One author is affiliated with a collagen research institute, which the paper discloses. There is no authorised claim for collagen, so this is a description of one trial rather than a recommendation.

Beyond supplements

Resistance and impact exercise is the non-supplement lever with trial support of its own; our exercise page reviews it with its sources. Protein supports the muscle that loads the bone: a consensus group recommends 1.0 to 1.2 g per kg of body weight a day for adults over 65, with more for those who exercise [8], and that range is commonly applied to midlife as well. Not smoking and keeping alcohol low both matter.

If you have risk factors for fracture, such as a parent who broke a hip, early menopause, long-term steroid use or a previous fracture, ask your GP about a fracture-risk assessment; they will decide whether a bone density scan is indicated. HRT has a substantial evidence base for bone in appropriate candidates, and that is a clinical conversation.

Where this leaves you

Reach the calcium and vitamin D intakes, from food first. Do not expect a supplement to do what loading exercise and enough protein do. Vitamin K2 and creatine on its own have been tested for bone density and did not move it. No food supplement may claim to act on osteoporosis, and none of the evidence above would support such a claim.

Published by: Suggestic Inc, herstack.co.uk. The HerStack editors. 2026-09-26. This page has not been reviewed by a clinician.