DHEA for Bone Mineral Density

DHEA (dehydroepiandrosterone) is an adrenal hormone that the body converts into estrogen and testosterone, and its blood levels slowly fall with age (Mayo Clinic). That raised the question of whether replacing it could slow the loss of bone mineral density (BMD), the measure on a DXA scan. The best answer so far is a qualified maybe for older women and a mostly no for older men. In a pooled analysis of four randomized trials, 295 women and 290 men aged 55 or older took DHEA or placebo for 12 months; women on DHEA gained 1.0% in lumbar spine BMD and 0.5% at the trochanter and held total hip BMD steady, while men had no BMD benefit (Jankowski 2019).

A 2019 meta-analysis points the same way: hip BMD rose in women taking DHEA but not in men, and its authors said more trials were needed before calling the effect clinically meaningful (Lin 2019). Several trials found no gain in their main analysis, no trial has measured fractures, and bone density is a stand-in for fracture risk rather than the same thing.

DHEA is a precursor hormone: the body uses it to make testosterone and estrogen (Mayo Clinic). In the pooled analysis, 50 mg a day raised blood levels of DHEA sulfate, estradiol and IGF-1 (insulin-like growth factor 1, a growth-promoting hormone) in both sexes and testosterone in women (Jankowski 2019); in the DAWN trial, testosterone, estradiol and IGF-1 rose in women but not in men (von Mühlen 2008). In that trial, a blood marker of bone breakdown (C-terminal telopeptide of type-1 collagen) fell in women but not in men (von Mühlen 2008).

In a randomized trial of 58 women and 61 men, the rise in hip BMD tracked the rise in estradiol, and adjusting for 12-month estradiol eliminated the DHEA effect. The authors concluded that the hip gains were mediated mainly by estradiol rather than by DHEA sulfate itself (Jankowski 2008).

Hormone levels and bone markers are steps along the way, not outcomes that matter to patients. Estradiol also rose in the men in these trials (Jankowski 2008), yet men showed no consistent BMD gain (Jankowski 2019), so a hormone rise does not by itself mean a bone benefit.

Genetic data add a different clue. In a Mendelian randomization analysis, which uses inherited variants as a stand-in for lifelong hormone levels, women genetically predicted to have higher DHEAS had higher lumbar spine BMD and a lower risk of forearm fracture (odds ratio 0.70, 95% CI 0.55-0.88 per standard deviation), with no clear link to hip fracture (Quester 2022). That concerns lifelong natural levels, not a supplement started in later life, so it does not show that taking DHEA prevents fractures.

Most trials in older adults gave 50 mg of DHEA by mouth once a day for 6 to 24 months, and several enrolled only people with low blood DHEA sulfate (Jankowski 2006; Villareal 2000; Nair 2006). Lupus trials used mostly 200 mg a day, with 100 mg in one open-label arm (Hartkamp 2004; Mease 2005; Sánchez-Guerrero 2008), and a small study in postmenopausal women taking long-term glucocorticoids used 25-50 mg of micronized DHEA (Papierska 2012). In one trial everyone also took vitamin D (16 mcg a day) and calcium (700 mg a day) (Weiss 2009).

Mayo Clinic notes that studies have often found the quality control of DHEA supplements to be low, so the amount on a label may not match the contents (Mayo Clinic). No trial has compared forms; a cream was tested in one small study only (Labrie 1997).

In 70 women and 70 men aged 60-88 with low DHEA sulfate, 50 mg a day for 12 months produced borderline hip gains in the intent-to-treat analysis (total hip 1.0%, P = 0.05; trochanter 1.2%, P = 0.06). In women only, lumbar spine BMD rose 2.2% (P = 0.04), though the sex-by-treatment interaction was borderline (P = 0.05) (Jankowski 2006). The 225-person DAWN trial in healthy adults aged 55-85 found a positive effect on lumbar spine BMD in women (P = 0.03), no effect at the hip, femoral neck or total body, and no change at any site in men (von Mühlen 2008).

In 113 adults aged 65-75 who all took calcium and vitamin D, women's spine BMD rose 1.7% in the randomized first year (P = 0.0003) and reached 3.6% above baseline after 2 years on DHEA (a within-group change: the placebo group rose 2.6% after switching to DHEA), while hip BMD did not change and men showed no difference (Weiss 2009). Year 2 was open-label, and bone markers fell in both groups in year 1. A 2-year trial in 87 men and 57 women found no relevant effect on body composition, strength or quality of life; men on DHEA had higher femoral neck BMD and women had higher BMD at the ultradistal radius (the wrist end of the forearm bone), with no major adverse effects (Nair 2006).

Not every trial found a benefit. In 99 frail women (mean age 76.6) with low DHEA sulfate and low bone mass, 50 mg a day for 6 months plus exercise, calcium and vitamin D produced no significant change in BMD or bone turnover markers, though lower-limb strength improved (Kenny 2010). A small 6-month trial in 36 adults with very low DHEA sulfate reported total body and lumbar spine gains of 1.6% and 2.5% against 18 untreated control subjects (Villareal 2000); it was small and short, its abstract does not mention a placebo or blinding, and its authors called it preliminary.

The pooled analysis of four trials found women on DHEA gained 1.0% (spine) and 0.5% (trochanter) and held total hip BMD, and men had no BMD benefit; its authors called DHEA a possible approach for preserving bone and muscle in women and asked whether the benefit-to-risk balance beats estrogen therapy (Jankowski 2019). The 2019 meta-analysis of randomized trials in healthy adults found hip BMD rose significantly in women (standardized mean difference -0.5, 95% CI -0.95 to -0.04, P = 0.03, favouring DHEA; the minus sign comes from placebo being the reference). That is a moderate effect with a wide interval. The trochanter result was borderline (P = 0.05), and similar results were not seen in men (Lin 2019).

In 55 women with lupus on long-term low-dose glucocorticoids, prasterone 200 mg a day for a year raised lumbar spine BMD 1.7% versus a 1.1% loss with placebo (P = 0.003) and total hip BMD 2.0% versus a 0.3% loss (P = 0.013) (Mease 2005). A larger trial in 155 women missed its 6-month primary endpoint: the lumbar spine difference was 0.003 versus -0.005 g/cm2 (P = 0.293). In its open-label extension the 200 mg dose gave a dose-dependent spine gain, 1.083% after 18 months (P = 0.042), with no change at the total hip (Sánchez-Guerrero 2008).

In 60 women with quiescent lupus taking 200 mg a day for a year, DHEA did not significantly change BMD in premenopausal women; BMD rose in postmenopausal women who were not taking bisphosphonates or estrogen, although the authors said small numbers preclude firmer conclusions (Hartkamp 2004). A 19-woman study of postmenopausal women on long-term glucocorticoids who could not take bisphosphonates found spine and femoral neck BMD rose after DHEA 25-50 mg was added, but it had no control group (Papierska 2012). For broader context, see DHEA for lupus.

In 106 people with Addison's disease, 50 mg a day of micronized DHEA for 12 months reversed ongoing BMD loss at the femoral neck (P below 0.05) but not at other sites (Gurnell 2008). See DHEA for Addison's disease.

In 61 young women with anorexia nervosa, DHEA 50 mg a day and hormone replacement each raised hip BMD about 1.7% over a year, but once weight gain was accounted for no treatment effect was detectable (Gordon 2002). Over 18 months in 80 females aged 13-27, DHEA plus a contraceptive pill preserved spine and whole-body BMD z scores while placebo lost them (Divasta 2012). A 2022 meta-analysis concluded DHEA alone does not improve BMD after adjusting for weight gain, that the combination helped only older adolescents and adults with closed growth plates, that one study suggested possible harm in younger adolescents with open growth plates, and that current evidence does not support DHEA (Lin 2022).

A 2020 Polish expert panel concluded that trials and meta-analyses indicate DHEA is effective in postmenopausal women with low BMD or osteoporosis, and no serious adverse effects have been reported (Rabijewski 2020). A 2022 review in the Journal of Clinical Endocrinology and Metabolism found no consistent benefit of DHEA for menopausal symptoms, sexual function, cognition or wellbeing in normal women, and noted that the risks of high physiologic or pharmacologic use include androgenic and estrogenic side effects, a concern for long-term use (Wierman 2022). Mayo Clinic describes the osteoporosis findings as mixed and rates DHEA overall as one to avoid (Mayo Clinic). The Polish statement is a consensus opinion rather than new trial data, and it sits against Mayo's reading and the mixed trial results above.

The Bone Health and Osteoporosis Foundation guide lists the FDA-approved options as bisphosphonates, estrogen-related therapy, parathyroid hormone analogs, denosumab, romosozumab and calcitonin; DHEA is not among them. It suggests considering medicine in postmenopausal women and men 50 and older with a T-score of -2.5 or lower, or with low bone mass and a 10-year hip fracture risk of 3% or more or major fracture risk of 20% or more (LeBoff 2022). For supplements with trial evidence on bone density, see vitamin D, calcium and collagen. For diagnosed osteoporosis, see our osteoporosis page.

DHEA is a hormone, and Mayo Clinic's overall rating for it is “avoid” because it can cause serious side effects. Mayo Clinic notes it might raise androgen levels and have a steroid effect, and that high doses or long-term use might raise the risk of hormone-sensitive cancers such as prostate and breast cancers, though more research is needed. Oily skin, acne and male-pattern hair growth in women can occur, DHEA might lower HDL (“good”) cholesterol, and it might worsen mood disorders and raise the risk of mania (Mayo Clinic).

In the Addison's disease trial, some older women reached above-normal DHEA sulfate and had mild androgenic side effects (Gurnell 2008), and the safety profile in the lupus trials reflected prasterone's weak androgenic properties (Sánchez-Guerrero 2008). The 2-year Nair trial reported no major adverse effects (Nair 2006). Safety beyond about 2 years and effects on fractures are untested, and one study in the 2022 meta-analysis suggested possible harm to BMD in younger adolescents with open growth plates (Lin 2022).

Mayo Clinic lists estrogen, testosterone, certain antidepressants and the sedative triazolam as possible interactions (Mayo Clinic). No trial has paired DHEA with a bisphosphonate, denosumab or another osteoporosis drug. The anorexia nervosa trials compared it with, or combined it with, an estrogen-progestin pill (Gordon 2002; Divasta 2012), and in the 60-woman lupus trial the BMD gain appeared only in postmenopausal women not already taking bisphosphonates or estrogen (Hartkamp 2004), so it is unknown whether DHEA adds anything for people already on treatment.