REF-1407 · Javed et al., 2019 · JAMA Network Open · Paper · PubMed · Free full text

Will hormone therapy keep my muscle after menopause?

Not on trial evidence: in 12 RCTs (4474 women, mean age 59) users lost 0.06 kg less lean mass than controls (95% CI −0.05 to 0.18) — no evidence of a difference; low certainty.

Main caveat: Low-certainty evidence. Half the trials at high risk of bias, imprecision in most subgroups, and publication bias by Egger and Begg tests (the authors note…
fairly strong evidenceMeasured on non-athletes and recreational people

Javed and colleagues asked whether oestrogen-based hormone therapy (HT), alone or with a progestogen, slows the loss of lean body mass in postmenopausal women, and whether the answer depends on HT type and dose, follow-up length, time since menopause, study quality or how lean mass was measured.

They searched MEDLINE, Embase, AgeLine, CINAHL and SportDiscus to 25 April 2018 for randomised trials in community-dwelling postmenopausal women aged 50 or older that compared HT with no HT or placebo and measured lean or fat-free mass by body scanning (DXA, dual-photon absorptiometry, bioelectrical impedance, MRI or CT).

Of 8961 records, 12 trials with 4474 recruited participants (mean age 59.0, SD 6.1) were analysed: 6 from the United States and 6 from Europe, 22 HT arms (15 oestrogen-progestogen, 7 oestrogen-only), median follow-up 2 years (6 months to 6 years).

Pooled with a fixed-effect model, HT users lost 0.06 kg less lean mass than controls (95% CI −0.05 to 0.18; P = .26; I² = 0%) — no evidence of a difference. Every subgroup stayed non-significant, from 0.06 kg more to 0.20 kg less.

Half the trials were at high risk of bias, funnel-plot asymmetry and the Egger and Begg tests suggested publication bias, and GRADE certainty was low. The authors judge any benefit unlikely to be clinically relevant for the average postmenopausal woman and conclude that interventions other than HT should be explored for muscle retention.

Our reading: this gives the base a direct answer to the athlete's question whether HRT will keep her muscle — on randomised evidence, no detectable effect on lean mass; it says nothing about strength (an older meta-analysis the authors cite found about 5% more strength on HT, not in the base) or about training response, and it is not a reason for or against HT, which is decided on symptoms and bone (REF-1406).

Protocol

Design: systematic review and meta-analysis of randomised clinical trials; PRISMA; registered in PROSPERO (CRD42016052047) on 30 November 2016.
Search: MEDLINE, Embase, AgeLine, CINAHL and SportDiscus from inception to 25 April 2018, plus reference lists; English-language, complete, published RCTs.
Eligibility: community-dwelling postmenopausal women aged 50 or older on oestrogen-based or oestrogen-progestogen HT vs no HT or placebo; hospital, long-term-care and disease-specific populations (cancer, Turner syndrome, anorexia) excluded; lean or fat-free mass measured by DXA, bioelectrical impedance, MRI, dual-photon absorptiometry or CT (skinfolds and circumferences excluded).
Included: n = 4474 recruited in 12 trials (8961 records → 219 → 21 full texts → 12; exclusions for measurement method, a cohort overlapping another Women's Health Initiative report, and missing numeric data).Age range across trials 45 to 75 years; ethnicity reported in 2 trials; physical activity reported in 3. 22 HT arms (15 oestrogen-progestogen, 7 oestrogen-only; 18 continuous, 4 cyclical); 21 arms pooled (one trial reported two arms combined). Doses standardised to conjugated equine oestrogen and split at 0.625 mg.
Outcome: absolute change in lean body mass (kg); lean mass by DXA in most trials, dual-photon absorptiometry or bioelectrical impedance in others.
Analysis: fixed-effect meta-analysis (Review Manager 5.3), mean difference with 95% CI and I²; subgroups by HT type and oestrogen dose, follow-up (2 years or less vs more), time since menopause (under 5 vs 5 years or more; and within 10 years), study quality and measurement type (DXA vs other). Cochrane risk-of-bias tool; GRADE; funnel plot with Egger and Begg tests (P < .10 = evidence of publication bias).
Funding and interests: no funding statement appears in the Europe PMC full text. Conflict of interest disclosures as printed: Dr Raina holds a Tier 1 Canada Research Chair in Geroscience and the Raymond and Margaret Labarge Chair in Research and Knowledge Application for Optimal Aging; no other disclosures were reported. No device maker, federation or data platform is named.

Key Results

AnalysisHT vs control, lean-mass change (kg; positive = less loss on HT)Certainty (GRADE)
All trials (21 arms)+0.06 (95% CI −0.05 to 0.18); P = .26; I² = 0%low
HT type and oestrogen dosefrom −0.06 to +0.19; none significantmoderate for oestrogen-only (any dose) and oestrogen-progestogen under 0.625 mg; low for oestrogen-progestogen at 0.625 mg or more
Follow-up (≤2 vs >2 years)0.0 to 0.10; none significantmoderate (shorter), low (longer)
Time since menopause (<5 vs ≥5 years)0.01 to 0.13; none significantlow for both
Study quality0.04 to 0.20; none significantmoderate (good), low (poor)
Measurement (DXA vs other)0.06 to 0.07; none significantmoderate (DXA), low (other)
Risk of bias: 6 of 12 trials (50%) high, 4 (33%) unclear, 2 (17%) low; mainly reporting deficiencies.
Publication bias: funnel-plot asymmetry; Egger P = .02, Begg P = .04 — smaller trials with larger effects may be over-represented, which would bias the pooled estimate toward a benefit of HT (the authors' point).
Single trials: 1 of 12 reached significance, a trial of 14 participants; 7 HT arms went with lean-mass loss and 14 with retention.
Size: the authors note that women over 50 lose about 1% of muscle mass a year and judge the pooled difference unlikely to be clinically relevant for the average postmenopausal woman.

Mechanisms

Measured: no detectable difference in whole-body lean mass between women on and off HT over a median of 2 years, at any dose, type, duration or time since menopause.
Authors' interpretation: oestrogen may act on muscle through receptors on skeletal muscle and indirectly through growth hormone and IGF-1, but loss of muscle after menopause is associated more with inactivity, low protein intake and oxidative stress than with menopause itself, which may be why HT shows no benefit; studies combining exercise and HT, in their reading, found no muscle-mass benefit over exercise alone.
Inference: whole-body lean mass by DXA or impedance includes water and organs and is a blunt measure of muscle; a small effect on muscle quality, strength or training response would not be visible in it. The base holds no trial of HT on strength, endurance or training adaptation in trained women.

Practical Application

Will hormone therapy keep my muscle after menopause? Not on trial evidence: in 12 RCTs (4474 women, mean age 59) users lost 0.06 kg less lean mass than controls (95% CI −0.05 to 0.18) — no evidence of a difference; low certainty.
Hormone therapy is not a muscle or performance tool on this evidence: the pooled difference was null across doses, types, durations and time since menopause (A-, SR/MA of RCTs, low certainty). Whether to use it is decided with a clinician, for symptoms and bone (REF-1406); the assistant states this finding and neither recommends nor discourages therapy.
The levers for muscle are training and food: in middle-aged women, resistance training raised strength whatever the menopause status, while muscle grew only before menopause at 6-8 sets per muscle a week and the authors infer that more volume is needed after it (REF-523, B+, non-athletes); heavier loads built more leg strength than lighter ones after 50 (REF-1086, A).Not from this paper; the base's own practice: protein near 1.8 g/kg spread through the day (SYNTH-69).
Strength is a separate question: the authors cite an older meta-analysis of 23 human studies (not in the base) in which HT went with about 5% more strength; this review did not test strength or function, so no effect on muscle mass is not the same as no effect on strength.
What it is good for as a source: a randomised-trial answer, with its certainty, to whether HT preserves lean mass in postmenopausal women. What it does not show: effects on strength, power, endurance, recovery or training adaptation; effects in athletes or in women with low muscle mass; anything about HT's benefits and risks for symptoms or bone.

Limitations

Low-certainty evidence. Half the trials at high risk of bias, imprecision in most subgroups, and publication bias by Egger and Begg tests (the authors note that their imprecision rule is hard to read when the true effect may be null).
Outcome. Whole-body lean mass, by three measurement methods (DXA, dual-photon absorptiometry, bioelectrical impedance), median follow-up 2 years; not muscle strength, function or performance.
Who. Postmenopausal women in general, mostly without reported physical activity; the window-of-opportunity analysis (HT within 10 years of menopause) was limited because the trials did not target very early menopause; no athletes; ethnicity reported in 2 trials.
Design details. Follow-up length was used as a proxy for duration of HT; cyclical vs continuous regimens, ethnicity and physical activity could not be analysed; several trials had unequal arm sizes that the authors say may indicate faulty randomisation or loss to follow-up.
Statements that do not add up: the Conclusions say "significant between study heterogeneity remained", while the Results report I² = 0% for the pooled analysis; the Results' first sentence speaks of 21 studies with 4474 participants where 21 is the number of full texts screened and 12 the number analysed; the Conclusions describe "individual study effects" ranging from −0.06 to 0.20 kg, which is the range the Results give for subgroups; and the inclusion rule was age 50 or older, but one included trial enrolled women aged 45-55.
Funding and interests: no funding statement in the full text; Dr Raina's research chairs are the only disclosures. No device maker, federation or data platform is named.
How this entry was checked and changed
Why this level

SR/MA of controlled trials, read from the full text (open access, CC-BY); held below A because the authors rate the overall certainty low (6 of 12 trials at high risk of bias, imprecision, Egger and Begg tests suggesting publication bias), half the trials recruited fewer than 100 women (our count from Table 1), lean mass was measured by three different methods, and the outcome is mass, not strength or function

Full text (checked 2026-10-28)
Of 8961 studies that met selection criteria, 12 were included, with a total of 4474 recruited participants; mean (SD) age was 59.0 (6.1) years (full text)
Of the 22 HT intervention arms, 15 used estrogen-progesterone combination HT and 7 used estrogen-only HT; the median follow-up duration was 2 years (range, 6 months to 6 years) (full text)
Across all studies, participants receiving HT lost 0.06 kg less LBM (95% CI, −0.05 to 0.18; I^2 = 0%) compared with those not receiving HT; P = .26 (full text)
HT users losing between 0.06 kg more to 0.20 kg less LBM compared with control participants for all strata. The quality of evidence based on GRADE was low (full text)
Six of the 12 studies (50%) were at high risk of bias, 4 (33%) were unclear, and 2 (17%) were at low risk of bias (full text)
The Egger and Begg tests also suggest publication bias (Egger P = .02; Begg P = .04) (full text)
Of the 12 included studies, 1 had a statistically significant result; however, it was severely limited by a sample size of 14 participants (full text)
It is reported that women older than 50 years lose approximately 1% of muscle mass annually (full text)
Dr Raina holds a Tier 1 Canada Research Chair in Geroscience and the Raymond and Margaret Labarge Chair in Research and Knowledge Application for Optimal Aging. No other disclosures were reported (full text)