There are 545 papers on GDF-15 in aging research: 34 trials and 342 observational studies in people, 48 in animals, 29 in cells, 85 reviews. The newest story, Nov 16, 2026: “Blood protein model better predicts future diabetes risk”.
“Similarly, plasma GDF-15 concentrations were 95% higher in the M+ than in the M– examination (Fig. 1C) and plasma glucagon levels were 26% higher (Fig. 1D), while serum cortisol levels and plasma lipids were similar between both conditions (Table 2).”
Metformin increases GDF-15 in high-fat diet-fed wild-type mice.
“metformin treatment ameliorated glucose intolerance and increased hepatic and circulating GDF15 levels in WT mice, but not in Ppard-/- mice fed a HFD”
High-intensity functional training decreases GDF-15 in men with metabolic syndrome.
“A significant difference was found between the two groups for FGF-21 [F(1, 21) = 97.54, P < 0.001, η2 = 0.852] and GDF-15 [F(1, 21) = 93.09, P < 0.001, η2 = 0.846].”
Human trialin men with metabolic syndrome12 weeksn = 24
Combined aerobic and resistance training increases GDF-15 in humans (group × time interaction across obesity groups) (group × time interaction p < 0.05; main group effect p < 0.05).
“led to increased GDF-15, indicating potential metabolic benefits”
Human trialgroup × time interaction across obesity groups16 weeksfemales
Nucleotides decrease GDF-15 in the Low-PRS group (β = −105.72, 95% CI −179.01 to −32.42).
“NTs supplementation also resulted in reductions of circulating GDF15 (β = −105.72, 95% CI −179.01 to −32.42, p = 0.0047) and IL‐1β (β = −92.27, 95% CI −177.71 to −6.84, p = 0.034).”
GDF-15 is linked with atrial fibrillation in humans (mediating the association of air pollution with AF risk).
“Mediation analysis identified specific pathogenic axes, highlighting ANGPT2, GDF15, and PLAUR as candidate molecular transducers linking pollution to AF risk.”
Humansmediating the association of air pollution with AF risk
GDF-15 was not clearly linked with cognitive decline in humans (direct effect, after accounting for physical frailty) (β = 0.100 (95% CI: -0.991 to 1.190)).
“whereas the direct effect was not significant (β = 0.100, 95% CI: -0.991 to 1.190)”
Humansdirect effect, after accounting for physical frailtyn = 134
GDF-15 is linked with all-cause dementia in humans (Mediation of the relationship between physical activity and incident all-cause dementia) (21 unique proteins identified as mediators).
“GDF15, ITGAV, ITGAM, ITGA11, HPGDS, GFAP, ADM, AHNAK, and DPP4 were among 21 unique proteins found to mediate the relationship of PA with all-cause dementia”
HumansMediation of the relationship between physical activity and incident all-cause dementiaplasma
GDF-15 is linked with all-cause dementia in humans (mediating the association between degenerative joint diseases and all-cause dementia) (mediation proportion 22%, 95% CI 14.5–46.8%).
“Cardiomyocyte-specific reduction of GDF15 expression preserves exercise tolerance and cardiac contractility in PGC-1α-deficient mice during endurance training.”
Animalsshown by cardiomyocyte-specific knockdown in PGC-1α-deficient mice during endurance trainingheart
GDF-15 is linked with lower appendicular lean mass in humans (Mendelian randomization using genetic instruments) (OR = 0.984, 95% CI = 0.974-0.993, pIVW = 7.658E-04).
“GDF-15 was negatively correlated with all four traits of sarcopenia”
HumansMendelian randomization using genetic instruments
GDF-15 is linked with dementia in humans (POHP-dementia association, as pure indirect effect) (explains 28% of the total effect of POHP on dementia).
“Growth differentiation factor 15 (GDF15) exhibited the strongest mediating effects (PIE > 0, P < 0.001), explaining 28% the total effect of POHP on dementia, as a pure indirect effect.”
HumansPOHP-dementia association, as pure indirect effectup to 15 years of follow-upn = 37,269
Prediction of non-cardiovascular and cancer mortality
↑upin humans1
1 study
GDF-15 is linked with better prediction of non-cardiovascular and cancer mortality in humans (Non-CV: cNRI 0.273, P=0.030; IDI 0.017, P=0.005. Cancer: cNRI 0.231, P=0.016; IDI 0.004, P=0.029).
“significantly improved the prediction of non-CV and cancer deaths in the entire cohort”
HumansAddition to a model containing clinical confounders and established cardiovascular biomarkers in the entire cohortserum6 yearsn = 1,281
GDF-15 showed no clear effect on HbA1c reduction in humans after adjusting for age and diabetes duration.
“no significant relationship was found between GDF15 (tertile 2: p = 0.403; tertile 3: p = 0.217) and ΔHbA1c after adjusting for age and diabetes duration”
Human trialafter adjusting for age and diabetes duration16 weeksn = 166
bioRxiv : the preprint server for biology·Preprint
Age is linked with higher GDF-15 in humans (1.1 ± 0.4 pg·mL−1 ·year−1 in women; 3.3 ± 0.6 pg·mL−1 ·year−1 in men).
“Systemic levels of GDF15 increased progressively as a function of age in women (1.1 ± 0.4 pg·mL−1 ·year−1 ) and men (3.3 ± 0.6 pg·mL−1 ·year−1 ) (both P < 0.05).”
age-multimorbidity ordinal category is linked with higher GDF-15 in humans (16.0% increase per category).
“In the hospital cohort, smoking and age-multimorbidity were independently associated with elevated GDF-15, with 26.1% and 16.0% increases per category, respectively (p < 0.05).”
Humanshospital cohort; multivariable linear regression adjusted for tobacco smoking and CD4+ nadir
“macrophage inhibitory cytokine 1 (MIC-1) (growth differentiation factor 15 (GDF15)), human epididymis protein 4 (HE4), and insulin-like growth factor-binding protein 2 (IGFBP-2) were associated with increased mortality”
GDF-15 is linked with a higher risk of mortality in humans (per SD, plasma proteomics) (HRperSD 1.56 (95% CI 1.42-1.72)).
“GDF15 (HRperSD [95% CI] = 1.56 [1.42-1.72]), WFDC2 (1.47 [1.33-1.62]) and TIMP1 (1.45 [1.29-1.62]) were the leading proteomic candidates associated with mortality.”
GDF-15 is associated with mortality in humans (normoglycemia and prediabetes).
“Three proteins (GDF15, EDA2R, and WFDC2) were shared across all groups, with GDF15 emerging as the top-ranked important protein in normoglycemia and prediabetes”
GDF-15 is associated with age in humans (plasma from the Baltimore Longitudinal Study of Aging (BLSA) cohort).
“Growth/differentiation factor 15 (GDF15), stanniocalcin 1 (STC1), and serine protease inhibitors (SERPINs), which significantly correlated with age in plasma from a human cohort”
Humansplasma from the Baltimore Longitudinal Study of Aging (BLSA) cohortplasma
GDF-15 is associated with sarcopenia in humans (adjusted multinomial logistic model; confirmed sarcopenia category) (aOR 1.9 (95% CI 1.3-2.6) for confirmed sarcopenia).
“higher GDF-15 was associated with greater odds of probable (aOR = 1.6; 95% CI: 1.2-2.0) and confirmed sarcopenia (aOR = 1.9; 95% CI: 1.3-2.6)”
GDF-15 is linked with a higher risk of sarcopenia in humans (particularly among individuals with low physical performance).
“elevated serum levels of GDF15 and its receptor GFRAL are associated with an increased risk of sarcopenia, particularly among individuals with low physical performance”
Humansparticularly among individuals with low physical performance
GDF-15 is associated with sarcopenia in humans (median 1916 vs 1569 pg/mL, p = 0.035).
“Median GDF-15 levels in patients with/without sarcopenia and with/without frailty were 1916 pg/mL vs. 1569 pg/mL (p = 0.035), and 2252 pg/mL vs. 1438 pg/mL (p < 0.001), respectively.”
GDF-15 is associated with sarcopenia in humans (acutely admitted older medical patients) (P < 0.01).
“Systemic GDF-15 was significantly higher in patients with either sarcopenia (P < 0.01) or frailty (P < 0.001) compared with patients without the conditions.”
Humansacutely admitted older medical patientsn = 1,036
GDF-15 is associated with sarcopenia in humans (highest vs lowest tertile, adjusted for cardiometabolic risk and biological factors) (OR 1.96 (95% CI 1.16-3.33)).
“the highest GDF-15 tertile (≥1245 pg/mL) had an increased likelihood of sarcopenia (odds ratio, 1.96; 95% confidence interval, 1.16-3.33)”
Humanshighest vs lowest tertile, adjusted for cardiometabolic risk and biological factorsn = 929
GDF-15 is associated with frailty in humans (cross-sectional comparison of high versus low plasma GDF-15) (OR 3.56 (95% CI 1.58-8.03), high vs low GDF-15).
“Higher plasma GDF-15 levels (both as continuous and categorical) were cross-sectionally associated with frailty (high vs. low GDF-15: OR = 3.56, 95% CI = 1.58-8.03).”
Humanscross-sectional comparison of high versus low plasma GDF-15
GDF-15 is linked with a higher risk of frailty in humans (baseline levels) (AUC 0.85 (Fried Phenotype), 0.96 (Clinical Frailty Scale)).
“Higher baseline GDF15 levels identified individuals at higher risk to become frail (AUC= 0.85 with Fried Phenotype and AUC= 0.96 with Clinical Frailty Scale)”
GDF-15 is associated with frailty in humans (acutely admitted older medical patients) (P < 0.001).
“Systemic GDF-15 was significantly higher in patients with either sarcopenia (P < 0.01) or frailty (P < 0.001) compared with patients without the conditions.”
Humansacutely admitted older medical patientsn = 1,036
GDF-15 is associated with frailty in humans (cross-sectional association with prevalent prefrailty and frailty) (p prefrailty = 1 × 10−15; p frailty = 2 × 10−19).
“higher odds of prefrailty and frailty were observed with higher levels of growth differentiation factor 15”
Humanscross-sectional association with prevalent prefrailty and frailtyn = 3,838
GDF-15 is associated with physical function in humans (grip strength or gait speed).
“Growth/differentiation factor 15, Insulin-like growth factor-binding protein 2, and Cystatin-C showed significant associations with inflammatory markers and measures of physical function (grip strength or gait speed).”
GDF-15 is associated with physical function in humans.
“Growth/differentiation factor 15, Insulin-like growth factor-binding protein 2, and Cystatin-C showed significant associations with inflammatory markers and measures of physical function”
GDF-15 is associated with muscle mass in men only, after adjustment for age and number of drugs per day.
“Elevated GDF15 concentrations were associated with lower measures of muscle mass, exclusively in men, after adjustment for age and number of drugs per day.”
Humansin men only, after adjustment for age and number of drugs per dayskeletal musclemales
GDF-15 is associated with telomere length in humans after adjustment for confounders, adjusted β = -0.200 (r = -0.434; adjusted β = -0.200 (95% CI -0.305 to -0.094)).
“negatively correlated with telomere length (r = -0.434, p < 0.001)”
Humansafter adjustment for confounders, adjusted β = -0.200PBMCmalesn = 120
GDF-15 is associated with body weight in the metformin group (r = −0.67).
“Changes in GDF-15 correlated inversely with changes in body weight in the metformin group (r = −0.67; P =.003), whereas no association was observed in the placebo group (r = 0.23; P =.359; and ).”
GDF-15 is associated with HbA1c reduction in humans (univariate model; tertile 3 versus lowest tertile reference) (β 0.553, 95% CI 0.115 to 0.991).
“Univariate linear regression analysis revealed a positive association between baseline GDF15 (tertile 3: β = 0.553, 95% CI 0.115 to 0.991, p = 0.014) and ΔHbA1c.”
Human trialunivariate model; tertile 3 versus lowest tertile reference16 weeksn = 166
Japan Society for the Promotion of Science¥32MGouvernement du Canada | Canadian Institutes of Health ResearchCA$17.5KNatural Science Foundation of Hunan Province
In about 20,000 UK Biobank participants, the model scored 0.86 for ranking diabetes risk, versus 0.74 for a blood-fat ratio.In about 20,000 UK Biobank participants, the model scored 0.86 for ranking diabetes risk, versus 0.74 for a blood-fat ratio.
Compared with participants without hypotension, those diagnosed with low blood pressure had a heart failure hazard ratio of 1.69 over a median 16.8 years.Compared with participants without hypotension, those diagnosed with low blood pressure had a heart failure hazard ratio of 1.69 over a median 16.8 years.