FitProven
Evidence Breakdown
Based on 9 studies
THE BIGGER QUESTION
Whether this one statement holds is settled above. What to actually do about it is a wider question, weighed across every claim that bears on it.
- Should women take creatine? — this claim is one of 3 weighed there
Evidence map
For & against, at a glance
Double-Blind RCT
Pro
Gualano B et al. · 2014Experimental Gerontology24-week randomized, double-blind, placebo-controlled trial in 60 vulnerable older women across four arms (placebo, creatine alone, resistance training alone, creatine + resistance training). Creatine + training produced the largest gain in appendicular lean mass (+1.31%, vs −1.2% placebo, +0.3% creatine alone and −0.2% training alone; p ≤ 0.05 against each) and was superior to every other group for 1-RM bench press (+10%). The leg-press result is narrower than it is often reported: creatine + training (+19.9%) beat placebo (+2.4%) and creatine alone (+3.7%), but **not** training alone (+15%, p = 0.357) — so on the biggest lower-body lift, adding creatine to training did not significantly outperform training by itself. Creatine without training was not inert either: its lean-mass gain was comparable to that of training alone (p = 0.62) and superior to placebo, though it was clearly the weakest of the two active strategies. Bone mass, serum bone markers and fat mass did not differ between groups — the benefit does not extend to bone.
0.55
24-week randomized, double-blind, placebo-controlled trial in 60 vulnerable older women across four arms (placebo, creatine alone, resistance training alone, creatine + resistance training). Creatine + training produced the largest gain in appendicular lean mass (+1.31%, vs −1.2% placebo, +0.3% creatine alone and −0.2% training alone; p ≤ 0.05 against each) and was superior to every other group for 1-RM bench press (+10%). The leg-press result is narrower than it is often reported: creatine + training (+19.9%) beat placebo (+2.4%) and creatine alone (+3.7%), but **not** training alone (+15%, p = 0.357) — so on the biggest lower-body lift, adding creatine to training did not significantly outperform training by itself. Creatine without training was not inert either: its lean-mass gain was comparable to that of training alone (p = 0.62) and superior to placebo, though it was clearly the weakest of the two active strategies. Bone mass, serum bone markers and fat mass did not differ between groups — the benefit does not extend to bone.
Design Double-Blind RCT (0.85) × quality 0.65 = impact 0.55
View sourceMeta-Analysis
Con
Delpino FM et al. · 2022NutritionSystematic review and meta-analysis of 35 RCTs (1,192 participants) on creatine and lean body mass, with age and sex subgroup analyses. Creatine plus resistance training added 1.10 kg of lean body mass (95% CI 0.56–1.65) and worked equally well across young, middle-aged and older adults — but the **sex subgroup went the opposite way to the claim**: males gained 1.46 kg (95% CI 0.47–2.46) while females gained a non-significant 0.29 kg (95% CI −0.43 to 1.01). Creatine without exercise did nothing (0.03 kg, 95% CI −0.65 to 0.70). Age is not the limiting factor, but women are, if anything, *less* responsive than men for lean mass — the opposite of "especially beneficial".
0.85
Systematic review and meta-analysis of 35 RCTs (1,192 participants) on creatine and lean body mass, with age and sex subgroup analyses. Creatine plus resistance training added 1.10 kg of lean body mass (95% CI 0.56–1.65) and worked equally well across young, middle-aged and older adults — but the **sex subgroup went the opposite way to the claim**: males gained 1.46 kg (95% CI 0.47–2.46) while females gained a non-significant 0.29 kg (95% CI −0.43 to 1.01). Creatine without exercise did nothing (0.03 kg, 95% CI −0.65 to 0.70). Age is not the limiting factor, but women are, if anything, *less* responsive than men for lean mass — the opposite of "especially beneficial".
Design Meta-Analysis (1.0) × quality 0.85 = impact 0.85
View sourceDouble-Blind RCT
Con
Chilibeck PD et al. · 2023Medicine and Science in Sports and ExerciseThe largest and longest trial in this population: 2-year randomized, double-blind, placebo-controlled study in 237 postmenopausal women (120 creatine at 0.14 g/kg/day, 117 placebo) with exercise. There was **no group difference in bone mineral density** — femoral neck (creatine 0.725 → 0.712 vs placebo 0.721 → 0.706 g/cm²), total hip, or lumbar spine (0.932 → 0.925 vs 0.923 → 0.915 g/cm²) — and strength (bench press and hack squat 1RM) improved similarly in both groups with no creatine advantage. The paper is not a clean null, though, and the record should not be read as one: creatine significantly maintained **section modulus** (p = 0.0011) and **buckling ratio** (p = 0.011) at the narrow part of the femoral neck, both geometric predictors of bending strength, and cut 80-m walking time (p = 0.0008); lean mass favored creatine in the completer subanalysis (p = 0.046). The authors' own conclusion is that two years of creatine plus exercise "had no effect on BMD; yet, it improved some bone geometric properties at the proximal femur." So the headline BMD case from the same group's earlier 12-month trial does not survive, but a weaker geometric signal does.
0.77
The largest and longest trial in this population: 2-year randomized, double-blind, placebo-controlled study in 237 postmenopausal women (120 creatine at 0.14 g/kg/day, 117 placebo) with exercise. There was **no group difference in bone mineral density** — femoral neck (creatine 0.725 → 0.712 vs placebo 0.721 → 0.706 g/cm²), total hip, or lumbar spine (0.932 → 0.925 vs 0.923 → 0.915 g/cm²) — and strength (bench press and hack squat 1RM) improved similarly in both groups with no creatine advantage. The paper is not a clean null, though, and the record should not be read as one: creatine significantly maintained **section modulus** (p = 0.0011) and **buckling ratio** (p = 0.011) at the narrow part of the femoral neck, both geometric predictors of bending strength, and cut 80-m walking time (p = 0.0008); lean mass favored creatine in the completer subanalysis (p = 0.046). The authors' own conclusion is that two years of creatine plus exercise "had no effect on BMD; yet, it improved some bone geometric properties at the proximal femur." So the headline BMD case from the same group's earlier 12-month trial does not survive, but a weaker geometric signal does.
Design Double-Blind RCT (0.85) × quality 0.90 = impact 0.77
View sourceMeta-Analysis
Con
Forbes SC et al. · 2018Frontiers in NutritionMeta-analysis of 5 RCTs (193 older adults, >50 years or postmenopausal) comparing creatine plus resistance training with resistance training alone. Pooled bone mineral density showed no benefit at any site: femoral neck (MD 0.00, 95% CI −0.01 to 0.01, p = 0.71), lumbar spine (MD 0.01, 95% CI −0.01 to 0.03, p = 0.32), hip (MD −0.01, p = 0.26) and whole body (MD 0.00, p = 0.50). Adding creatine to training does not improve BMD over training alone.
0.60
Meta-analysis of 5 RCTs (193 older adults, >50 years or postmenopausal) comparing creatine plus resistance training with resistance training alone. Pooled bone mineral density showed no benefit at any site: femoral neck (MD 0.00, 95% CI −0.01 to 0.01, p = 0.71), lumbar spine (MD 0.01, 95% CI −0.01 to 0.03, p = 0.32), hip (MD −0.01, p = 0.26) and whole body (MD 0.00, p = 0.50). Adding creatine to training does not improve BMD over training alone.
Design Meta-Analysis (1.0) × quality 0.60 = impact 0.60
View sourceDouble-Blind RCT
Pro
Gualano B et al. · 2014Experimental Gerontology24-week randomized, double-blind, placebo-controlled trial in 60 vulnerable older women across four arms (placebo, creatine alone, resistance training alone, creatine + resistance training). Creatine + training produced the largest gain in appendicular lean mass (+1.31%, vs −1.2% placebo, +0.3% creatine alone and −0.2% training alone; p ≤ 0.05 against each) and was superior to every other group for 1-RM bench press (+10%). The leg-press result is narrower than it is often reported: creatine + training (+19.9%) beat placebo (+2.4%) and creatine alone (+3.7%), but **not** training alone (+15%, p = 0.357) — so on the biggest lower-body lift, adding creatine to training did not significantly outperform training by itself. Creatine without training was not inert either: its lean-mass gain was comparable to that of training alone (p = 0.62) and superior to placebo, though it was clearly the weakest of the two active strategies. Bone mass, serum bone markers and fat mass did not differ between groups — the benefit does not extend to bone.
0.55
24-week randomized, double-blind, placebo-controlled trial in 60 vulnerable older women across four arms (placebo, creatine alone, resistance training alone, creatine + resistance training). Creatine + training produced the largest gain in appendicular lean mass (+1.31%, vs −1.2% placebo, +0.3% creatine alone and −0.2% training alone; p ≤ 0.05 against each) and was superior to every other group for 1-RM bench press (+10%). The leg-press result is narrower than it is often reported: creatine + training (+19.9%) beat placebo (+2.4%) and creatine alone (+3.7%), but **not** training alone (+15%, p = 0.357) — so on the biggest lower-body lift, adding creatine to training did not significantly outperform training by itself. Creatine without training was not inert either: its lean-mass gain was comparable to that of training alone (p = 0.62) and superior to placebo, though it was clearly the weakest of the two active strategies. Bone mass, serum bone markers and fat mass did not differ between groups — the benefit does not extend to bone.
Design Double-Blind RCT (0.85) × quality 0.65 = impact 0.55
View sourceMeta-Analysis
Con
Delpino FM et al. · 2022NutritionSystematic review and meta-analysis of 35 RCTs (1,192 participants) on creatine and lean body mass, with age and sex subgroup analyses. Creatine plus resistance training added 1.10 kg of lean body mass (95% CI 0.56–1.65) and worked equally well across young, middle-aged and older adults — but the **sex subgroup went the opposite way to the claim**: males gained 1.46 kg (95% CI 0.47–2.46) while females gained a non-significant 0.29 kg (95% CI −0.43 to 1.01). Creatine without exercise did nothing (0.03 kg, 95% CI −0.65 to 0.70). Age is not the limiting factor, but women are, if anything, *less* responsive than men for lean mass — the opposite of "especially beneficial".
0.85
Systematic review and meta-analysis of 35 RCTs (1,192 participants) on creatine and lean body mass, with age and sex subgroup analyses. Creatine plus resistance training added 1.10 kg of lean body mass (95% CI 0.56–1.65) and worked equally well across young, middle-aged and older adults — but the **sex subgroup went the opposite way to the claim**: males gained 1.46 kg (95% CI 0.47–2.46) while females gained a non-significant 0.29 kg (95% CI −0.43 to 1.01). Creatine without exercise did nothing (0.03 kg, 95% CI −0.65 to 0.70). Age is not the limiting factor, but women are, if anything, *less* responsive than men for lean mass — the opposite of "especially beneficial".
Design Meta-Analysis (1.0) × quality 0.85 = impact 0.85
View sourceDouble-Blind RCT
Con
Chilibeck PD et al. · 2023Medicine and Science in Sports and ExerciseThe largest and longest trial in this population: 2-year randomized, double-blind, placebo-controlled study in 237 postmenopausal women (120 creatine at 0.14 g/kg/day, 117 placebo) with exercise. There was **no group difference in bone mineral density** — femoral neck (creatine 0.725 → 0.712 vs placebo 0.721 → 0.706 g/cm²), total hip, or lumbar spine (0.932 → 0.925 vs 0.923 → 0.915 g/cm²) — and strength (bench press and hack squat 1RM) improved similarly in both groups with no creatine advantage. The paper is not a clean null, though, and the record should not be read as one: creatine significantly maintained **section modulus** (p = 0.0011) and **buckling ratio** (p = 0.011) at the narrow part of the femoral neck, both geometric predictors of bending strength, and cut 80-m walking time (p = 0.0008); lean mass favored creatine in the completer subanalysis (p = 0.046). The authors' own conclusion is that two years of creatine plus exercise "had no effect on BMD; yet, it improved some bone geometric properties at the proximal femur." So the headline BMD case from the same group's earlier 12-month trial does not survive, but a weaker geometric signal does.
0.77
The largest and longest trial in this population: 2-year randomized, double-blind, placebo-controlled study in 237 postmenopausal women (120 creatine at 0.14 g/kg/day, 117 placebo) with exercise. There was **no group difference in bone mineral density** — femoral neck (creatine 0.725 → 0.712 vs placebo 0.721 → 0.706 g/cm²), total hip, or lumbar spine (0.932 → 0.925 vs 0.923 → 0.915 g/cm²) — and strength (bench press and hack squat 1RM) improved similarly in both groups with no creatine advantage. The paper is not a clean null, though, and the record should not be read as one: creatine significantly maintained **section modulus** (p = 0.0011) and **buckling ratio** (p = 0.011) at the narrow part of the femoral neck, both geometric predictors of bending strength, and cut 80-m walking time (p = 0.0008); lean mass favored creatine in the completer subanalysis (p = 0.046). The authors' own conclusion is that two years of creatine plus exercise "had no effect on BMD; yet, it improved some bone geometric properties at the proximal femur." So the headline BMD case from the same group's earlier 12-month trial does not survive, but a weaker geometric signal does.
Design Double-Blind RCT (0.85) × quality 0.90 = impact 0.77
View sourceMeta-Analysis
Con
Forbes SC et al. · 2018Frontiers in NutritionMeta-analysis of 5 RCTs (193 older adults, >50 years or postmenopausal) comparing creatine plus resistance training with resistance training alone. Pooled bone mineral density showed no benefit at any site: femoral neck (MD 0.00, 95% CI −0.01 to 0.01, p = 0.71), lumbar spine (MD 0.01, 95% CI −0.01 to 0.03, p = 0.32), hip (MD −0.01, p = 0.26) and whole body (MD 0.00, p = 0.50). Adding creatine to training does not improve BMD over training alone.
0.60
Meta-analysis of 5 RCTs (193 older adults, >50 years or postmenopausal) comparing creatine plus resistance training with resistance training alone. Pooled bone mineral density showed no benefit at any site: femoral neck (MD 0.00, 95% CI −0.01 to 0.01, p = 0.71), lumbar spine (MD 0.01, 95% CI −0.01 to 0.03, p = 0.32), hip (MD −0.01, p = 0.26) and whole body (MD 0.00, p = 0.50). Adding creatine to training does not improve BMD over training alone.
Design Meta-Analysis (1.0) × quality 0.60 = impact 0.60
View sourceShowing the 4 strongest of 9 studies. Tap any node to expand its detail.
Evidence
PRO (4)
PRO Double-Blind RCTn=600.65 Gualano B, Macedo AR et al. (2014)
Creatine + training raised appendicular lean mass +1.31% (vs -1.2% placebo, +0.3% creatine alone, -0.2% training alone; all p <= 0.05) and was superior to every group for 1RM bench press (+10%). For 1RM leg press, creatine + training (+19.9%) beat placebo (+2.4%) and creatine alone (+3.7%) but NOT training alone (+15%, p = 0.357). Creatine alone gained lean mass comparably to training alone (p = 0.62) and more than placebo. No effect on bone mass, bone markers or fat mass
24-week randomized, double-blind, placebo-controlled trial in 60 vulnerable older women across four arms (placebo, creatine alone, resistance training alone, creatine + resistance training). Creatine + training produced the largest gain in appendicular lean mass (+1.31%, vs −1.2% placebo, +0.3% creatine alone and −0.2% training alone; p ≤ 0.05 against each) and was superior to every other group for 1-RM bench press (+10%). The leg-press result is narrower than it is often reported: creatine + training (+19.9%) beat placebo (+2.4%) and creatine alone (+3.7%), but not training alone (+15%, p = 0.357) — so on the biggest lower-body lift, adding creatine to training did not significantly outperform training by itself. Creatine without training was not inert either: its lean-mass gain was comparable to that of training alone (p = 0.62) and superior to placebo, though it was clearly the weakest of the two active strategies. Bone mass, serum bone markers and fat mass did not differ between groups — the benefit does not extend to bone.
Weighted 0.65 — A 24-week double-blind four-arm trial, but only 60 women were randomised (about 15 per arm) and completion numbers are not reported in any accessible copy. The funding statement could not be retrieved, so independence is unverified. The four-arm design is the real strength - it isolates creatine from training.
Experimental Gerontology
PRO Meta-Analysisn=1760.55 Dos Santos EEP, de Araújo RC et al. (2021)
Upper-body strength improved with creatine; no effect on lower-body strength or on any measure of muscle mass
Systematic review and meta-analysis of 10 RCTs (8 pooled; 211 participants, 176 in the meta-analysis) in older females aged 56–70, including healthy, postmenopausal and osteopenic/osteoarthritic women, comparing creatine plus resistance training against resistance training alone. Upper-body strength improved significantly with creatine, but there was no effect on lower-body strength or on any measure of muscle mass when all studies were pooled; sub-analysis suggested benefits emerged only in interventions of 24 weeks or longer. GRADE certainty was rated low because of the small total sample. Supports the muscle/strength half of the claim, but far more weakly than the marketing implies.
Weighted 0.55 — 176 participants pooled across 8 RCTs (10 RCTs, 211 participants in the review) - a small evidence base. The authors rate GRADE certainty low, benefits emerged only in interventions of 24 weeks or longer, and two authors sit on a creatine manufacturer's advisory board, so the positive strength finding warrants caution.
Funding: none declared; authors Candow and Forbes disclose creatine-industry ties (Alzchem scientific advisory board)
Nutrients
PRO Double-Blind RCTn=330.45 Chilibeck PD, Candow DG et al. (2015)
Femoral neck BMD declined less on creatine (-1.2%) than placebo (-3.9%, p < 0.05); femoral shaft subperiosteal width maintained (+0.04 cm vs -0.12 cm); relative bench press +64% vs +34%
12-month randomized, double-blind trial in 47 postmenopausal women (mean age 57; 33 completers) doing supervised resistance training 3 d/wk with 0.1 g/kg/day creatine or placebo. Femoral neck BMD declined less on creatine (−1.2%) than on placebo (−3.9%, p < 0.05), femoral shaft subperiosteal width was maintained rather than lost (+0.04 cm vs −0.12 cm, p < 0.05), and relative bench press strength rose more on creatine (+64% vs +34%). Note this is BMD preservation, not gain, and the completer sample was small — a larger, longer replication (Chilibeck 2023) did not reproduce the BMD finding.
Weighted 0.45 — 33 of 47 randomised women completed 12 months — 30% attrition, and the completer sample is small for a BMD endpoint. The paper is paywalled and no funding or COI statement could be verified, so whether the creatine was industry-donated is unknown. Decisively, the same group's larger and longer 2-year trial (chilibeck-2023, n=237) did not reproduce the BMD finding, which marks this as a probable small-study effect.
Medicine and Science in Sports and Exercise
PRO Narrative ReviewCONFLICTED0.20 Smith-Ryan AE, Cabre HE et al. (2021)
Argues women have 70-80% lower endogenous creatine stores than men and that postmenopausal women may gain skeletal muscle size and function at high doses (0.3 g/kg/day -- a 5-day LOADING dose, far above the 3-5 g/day maintenance dose typically sold). ON BONE THE PAPER CONTRADICTS ITSELF: the abstract claims 'favorable effects on bone when combined with resistance training', but the body text (section 5.2) concludes 'Creatine supplementation alone or in combination with resistance training appears to provide no benefits in bone physiology in post-menopausal females.'
Narrative review laying out the mechanistic case for creatine across the female lifespan: "females exhibiting 70-80% lower endogenous creatine stores compared to males," and the fall in estrogen at menopause affects creatine kinetics, muscle and bone. The authors conclude that "post-menopausal females may also experience benefits in skeletal muscle size and function when consuming high doses of creatine (0.3 g·kg⁻¹·d⁻¹)." Note what that dose is: 0.3 g/kg/day is the 5-day loading protocol (~20 g/day), not a maintenance dose — the paper's own maintenance figure is 3–5 g/day (0.03 g/kg/day), an order of magnitude lower.
On bone, the paper contradicts itself and the optimistic half is the one that gets quoted. The abstract asserts "favorable effects on bone when combined with resistance training." But the body text (section 5.2, Combined with Resistance Training) states: "Creatine supplementation alone or in combination with resistance training appears to provide no benefits in bone physiology in post-menopausal females." The body acknowledges mixed results across studies — some showed reduced hip bone-density loss and increased femoral shaft width, others "failed to show greater effects from CrM supplementation…on measures of bone mineral" — and lands on no consistent benefit. Any claim that this paper supports creatine for bone in postmenopausal women is resting on its abstract while its own analysis says otherwise.
It is the origin of much of the "creatine is for women over 40" messaging, but it is a review by the field's advocates, not primary evidence, and its own recommendations are more conditional than the messaging that followed.
Weighted 0.20 — No sample size: non-systematic narrative review, so study selection is unconstrained. Serious independence problem: the publishing fee was paid by Alzchem, a creatine manufacturer, and 2 of the 4 authors serve on Alzchem's scientific advisory board. An industry-funded review arguing FOR the funder's product is the weakest configuration of evidence, and this paper is the origin of much of the 'creatine is for women over 40' marketing while its own recommendations are more conditional than that messaging. Weighted near the floor of its class.
Funding: Alzchem (creatine manufacturer; paid the publishing fee) — the funder profits from this result.
Nutrients
AGAINST (3)
AGAINST Double-Blind RCTn=2370.90 Chilibeck PD, Candow DG et al. (2023)
NO group difference over 2 years in the primary outcome, femoral neck BMD (0.725 to 0.712 vs 0.721 to 0.706 g/cm2), nor in total hip or lumbar spine BMD; strength (bench press, hack squat 1RM) improved equally in both groups. Creatine DID maintain two proximal-femur geometric properties - section modulus (p = 0.0011) and buckling ratio (p = 0.011) - and reduced 80-m walking time (p = 0.0008); lean mass favoured creatine in the completer subanalysis (p = 0.046). The authors' conclusion is no effect on BMD, but improvement in some bone geometric properties
The largest and longest trial in this population: 2-year randomized, double-blind, placebo-controlled study in 237 postmenopausal women (120 creatine at 0.14 g/kg/day, 117 placebo) with exercise. There was no group difference in bone mineral density — femoral neck (creatine 0.725 → 0.712 vs placebo 0.721 → 0.706 g/cm²), total hip, or lumbar spine (0.932 → 0.925 vs 0.923 → 0.915 g/cm²) — and strength (bench press and hack squat 1RM) improved similarly in both groups with no creatine advantage. The paper is not a clean null, though, and the record should not be read as one: creatine significantly maintained section modulus (p = 0.0011) and buckling ratio (p = 0.011) at the narrow part of the femoral neck, both geometric predictors of bending strength, and cut 80-m walking time (p = 0.0008); lean mass favored creatine in the completer subanalysis (p = 0.046). The authors' own conclusion is that two years of creatine plus exercise "had no effect on BMD; yet, it improved some bone geometric properties at the proximal femur." So the headline BMD case from the same group's earlier 12-month trial does not survive, but a weaker geometric signal does.
Weighted 0.90 — 237 randomised and analysed by intention-to-treat over 2 years — the largest and longest trial in this population, publicly funded, double-blind, with a pre-specified primary outcome. An author sits on AlzChem's scientific advisory board and One author sits on AlzChem's scientific advisory board (a personal conflict, not study funding).
Funding: Canadian Institutes of Health Research (funding reference number 130235) and the Canada Foundation for Innovation
Medicine and Science in Sports and Exercise
AGAINST Meta-Analysisn=11920.85 Delpino FM, Figueiredo LM et al. (2022)
Creatine plus resistance training added 1.10 kg lean body mass (95% CI 0.56-1.65), equally across age groups; males gained 1.46 kg (95% CI 0.47-2.46) vs a non-significant 0.29 kg in females (95% CI -0.43 to 1.01); creatine without exercise did nothing (0.03 kg)
Systematic review and meta-analysis of 35 RCTs (1,192 participants) on creatine and lean body mass, with age and sex subgroup analyses. Creatine plus resistance training added 1.10 kg of lean body mass (95% CI 0.56–1.65) and worked equally well across young, middle-aged and older adults — but the sex subgroup went the opposite way to the claim: males gained 1.46 kg (95% CI 0.47–2.46) while females gained a non-significant 0.29 kg (95% CI −0.43 to 1.01). Creatine without exercise did nothing (0.03 kg, 95% CI −0.65 to 0.70). Age is not the limiting factor, but women are, if anything, less responsive than men for lean mass — the opposite of "especially beneficial".
Weighted 0.85 — 35 RCTs pooling 1,192 participants with pre-specified age and sex subgroups — large and publicly funded. Two reasons it is not top-band: the sex finding rests on subgroup analysis, where the female estimate is imprecise and a null result may reflect fewer and smaller female trials rather than a true absence of effect; and two co-authors have documented creatine-industry ties (advisory board, research donations) that could not be confirmed or excluded from this specific paper because it is paywalled.
Funding: CNPq and CAPES (Brazilian federal research agencies)
Nutrition
AGAINST Meta-Analysisn=1930.60 Forbes SC, Chilibeck PD et al. (2018)
No BMD benefit at any site: femoral neck MD 0.00 (95% CI -0.01 to 0.01), lumbar spine MD 0.01, hip MD -0.01, whole body MD 0.00
Meta-analysis of 5 RCTs (193 older adults, >50 years or postmenopausal) comparing creatine plus resistance training with resistance training alone. Pooled bone mineral density showed no benefit at any site: femoral neck (MD 0.00, 95% CI −0.01 to 0.01, p = 0.71), lumbar spine (MD 0.01, 95% CI −0.01 to 0.03, p = 0.32), hip (MD −0.01, p = 0.26) and whole body (MD 0.00, p = 0.50). Adding creatine to training does not improve BMD over training alone.
Weighted 0.60 — 193 participants across only 5 RCTs - a thin base for a meta-analysis, though the null is consistent across all four skeletal sites and the intervals are tight enough to exclude a meaningful BMD effect. The same author group discloses creatine-industry ties elsewhere but declares no commercial relationships here.
Frontiers in Nutrition
NEUTRAL (2)
NEUTRAL Meta-Analysisn=2250.40 Prokopidis K, Giannos P et al. (2023)
Small pooled benefit on memory (SMD 0.29, 95% CI 0.04-0.53, p = 0.02), driven entirely by adults aged 66-76 (SMD 0.88, p = 0.009), with nothing in 11-31 year-olds (SMD 0.03) and a non-significant female-only subgroup (SMD 0.39, p = 0.10)
Meta-analysis of 8 RCTs (225 healthy participants; 151 female) on creatine and memory. The pooled effect was small but significant (SMD 0.29, 95% CI 0.04–0.53, p = 0.02) and was driven entirely by older adults aged 66–76 (SMD 0.88, p = 0.009) with nothing in 11–31 year-olds (SMD 0.03, p = 0.72). Crucially for this claim, the female-only subgroup was not significant (SMD 0.39, p = 0.10), heterogeneity was high (I² = 66%), and a published Letter to the Editor argued the analysis double-counted arms and produced false-positive findings. The cognitive promise is the weakest leg of the claim.
Weighted 0.40 — 8 RCTs and only 225 participants — a thin pooled sample split further into age subgroups, with high heterogeneity (I² = 66%) and a published Letter to the Editor arguing the analysis double-counted arms and generated false positives. Combined with two authors' creatine-industry ties on a positive result, this is the weakest meta-analysis in this set.
Funding: No external funding; two authors declare creatine-industry ties (scientific advisory board of a creatine manufacturer, industry-sponsored creatine research, product donations and conference travel support)
Nutrition Reviews
NEUTRAL Systematic Reviewn=15420.35 Marshall S, Kitzan A et al. (2026)
5 of 6 studies reported positive associations, mostly in memory and attention, but the authors conclude only that creatine may be associated with cognitive benefit
Systematic review of creatine and cognition in adults aged 55+, covering 6 studies and 1,542 participants (55.7% female). Five of the six reported positive associations, mostly in memory and attention, but the authors rated only one study as "good" quality and three as "poor", and concluded only that creatine "may be associated with benefits for cognition" pending higher-quality trials. Genuinely inconclusive: the direction is favorable but the evidence base is too thin and too low-quality to support the confident cognitive claims made in creatine marketing to midlife women.
Weighted 0.35 — k=6 studies and 1,542 participants, but the total is dominated by a single large study (individual n ranged from 27 to 1,340), and the authors rated only one study "good" while three were "poor". No meta-analysis was performed, several designs were cross-sectional, and half the studies lacked a sample-size calculation. Independently funded and appropriately cautious in its own conclusions; the weight is low because the underlying evidence base is thin and poor-quality, not because the review is careless.
Funding: Natural Sciences and Engineering Research Council of Canada and the Canada Research Chairs Program
Nutrition Reviews