Handgrip Strength as an Aging Biomarker

Handgrip strength (HGS) is the single most validated physical-performance biomarker of biological aging in humans. It integrates skeletal muscle mass, neural drive, and peripheral circulation into one inexpensive, reproducible measure that consistently predicts all-cause mortality, cardiovascular mortality, disability, frailty, and sarcopenia across populations spanning five continents. The landmark PURE study (n=139,691; 17 countries) found HGS to be a stronger predictor of all-cause mortality than systolic blood pressure ¹. HGS requires no laboratory, no biospecimen, and a ~$60–300 dynamometer — placing it in a unique tier among aging biomarkers for accessibility and clinical deployability.

This page covers HGS as a quantitative aging biomarker: measurement protocol, aging trajectory, major mortality/sarcopenia cohorts, EWGSOP2 diagnostic cutoffs, Mendelian randomization status, intervention responsiveness, and hallmark linkages. For the underlying muscle biology, see sarcopenia.

Identity

Gold-standard instrument: JAMAR hydraulic dynamometer (Sammons Preston, Rolyan) — the most widely validated device; recommended in EWGSOP2 and most clinical guidelines ²
Alternative: Smedley spring-based dynamometer — less expensive; slightly lower values than JAMAR; results are not interchangeable without platform-specific norms
Unit: kilograms-force (kg); some devices report Newtons — convert N ÷ 9.81 for kg
Pre-analytical considerations: Dominant vs. non-dominant hand; position (seated, elbow at 90°, wrist neutral); 3 trials with 1–2-minute rest between; record maximum value; avoid testing immediately post-exercise

Standard protocol (EWGSOP2 / Southampton)

Patient seated, elbow flexed at 90°, forearm in neutral rotation
Use dominant hand unless contraindicated (arthritis, recent injury)
Three maximal squeezes, with verbal encouragement; 1-minute rest between trials
Record maximum of three values
Compare against age- and sex-specific normative reference data

The Southampton protocol (Roberts 2011) is the most widely cited standardization reference. Deviation from protocol (e.g., standing position, arm at side) can raise values ~10–15% — making protocol fidelity essential for longitudinal tracking.

Aging Trajectory

HGS peaks at approximately up to age ~40 in both sexes (young adulthood), with men averaging 45–55 kg and women 28–35 kg at peak ². The EWGSOP2 consensus paper (Cruz-Jentoft 2019) cites loss of muscle strength at 1.5–5% per year after age 50, accelerating after age 70 ². The Granic 2016 Newcastle 85+ longitudinal study quantified decline in the very old as −1.13 kg/year in men (linear) and a curvilinear pattern with acceleration in women ³.

Key age-related drivers of HGS decline:

Loss of muscle fiber cross-sectional area (especially type II fast-twitch)
Reduced alpha-motor neuron firing rate and neuromuscular junction efficiency
Mitochondrial dysfunction reducing ATP production in muscle fibers (see mitochondrial-dysfunction)
Satellite cell depletion reducing regenerative capacity (see stem-cell-exhaustion)
Chronic low-grade inflammation suppressing protein synthesis (see chronic-inflammation)

Approximate age-stratified means (healthy community-dwelling; dominant hand)

The table below synthesizes data from Massy-Westropp 2011 and Bohannon 2019 NHANES. Values are approximate summaries across populations; population-specific norms (European, US, Asian) differ meaningfully. needs-replication — no single global normative reference has been universally adopted.

Age group	Men (kg)	Women (kg)	Notes
20–29 yrs	~45–52	~28–33	Peak decade in most populations
30–39 yrs	~45–54	~28–34	Peak sustained; slight increase possible
40–49 yrs	~43–50	~27–32	Plateau or early decline begins
50–59 yrs	~39–46	~25–30	Meaningful decline; begins to distinguish fit vs. sedentary
60–69 yrs	~34–41	~22–27	Community-dwelling norms; EWGSOP2 cut-off zone
70–79 yrs	~28–35	~18–23	High proportion approaching EWGSOP2 threshold
80–89 yrs	~22–28	~14–19	Many below diagnostic threshold; sarcopenia prevalent
90+ yrs	~16–22	~10–16	Centenarian survivors tend to preserve relative strength

Sarcopenia Diagnostic Role: EWGSOP2 Cutoffs

The European Working Group on Sarcopenia in Older People 2 (EWGSOP2) consensus (Cruz-Jentoft 2019) uses low grip strength as the primary criterion for probable sarcopenia — the first step in the Find-Assess-Confirm-Severity (F-A-C-S) algorithm ².

EWGSOP2 cutoffs for low muscle strength (Table 3 of Cruz-Jentoft 2019, confirmed against local PDF):

Sex	HGS threshold	Severity level
Men	< 27 kg	Low strength (probable sarcopenia)
Women	< 16 kg	Low strength (probable sarcopenia)

These cutoffs are set at approximately −2 standard deviations from young-adult reference values (T-score ≤ −2.5 in Figure 2 of the paper). They are consistent with a Dodds 2014 normative dataset of UK adults and have been adopted by major international sarcopenia guidelines.

Diagnostic algorithm (EWGSOP2 F-A-C-S):

Find: SARC-F questionnaire or clinical suspicion (falls, weakness, slow gait)
Assess: HGS or chair-stand test — if low → probable sarcopenia
Confirm: DXA or BIA to document low muscle quantity/quality
Severity: Add gait speed, SPPB, TUG, or 400-m walk — if all three criteria met → severe sarcopenia

HGS is also a component of the Fried Frailty Phenotype (criterion: grip strength in lowest 20th percentile by sex and BMI quartile), creating direct overlap between the sarcopenia and frailty diagnostic frameworks.

Major Mortality Cohorts

Leong 2015 — PURE Study (Lancet)

The landmark Prospective Urban Rural Epidemiology study enrolled 139,691 participants aged 35–70 from 17 countries (low-, middle-, and high-income) across four continents; median follow-up 4.0 years ¹.

Key findings (from PubMed abstract; full PDF pending verification):

Each 5 kg decrease in HGS associated with HR 1.16 (95% CI 1.13–1.20) for all-cause mortality
Cardiovascular mortality HR: 1.17 (95% CI 1.11–1.24) per 5-kg decrement
Grip strength was a stronger predictor of all-cause and cardiovascular mortality than systolic blood pressure — a finding that shifted clinical perception of physical performance as a vital sign
The association held across high-, middle-, and low-income countries, establishing HGS as a globally applicable biomarker
No significant association with diabetes, falls, or fractures in this cohort (negative finding)

Dimension	Status
Pathway conserved in humans?	yes — direct human cohort
Phenotype conserved across populations?	yes — 17 countries, all income levels
Replicated in humans?	yes — multiple independent cohorts

Celis-Morales 2018 — UK Biobank (BMJ)

Among 502,293 UK Biobank participants aged 40–69 years (mean follow-up 7.1 years, range 5.3–9.9 yr), lower HGS strongly predicted mortality and disease incidence across multiple endpoints ⁴:

Outcome	HR per 5-kg lower HGS (women)	HR per 5-kg lower HGS (men)
All-cause mortality	1.20 (95% CI 1.17–1.23)	1.16 (1.15–1.17)
Cardiovascular mortality	1.19 (1.13–1.25)	1.22 (1.18–1.26)
Cancer (all sites)	not reported separately	—

Addition of HGS measurement enhanced cardiovascular risk prediction models beyond conventional clinical risk factors. The sample size (~3.6× PURE) and breadth of outcome ascertainment make this the largest single-cohort HGS mortality dataset.

Ling 2010 — Leiden 85-plus Study

In 555 participants at age 85 (65% women), followed up to age 94 (9.5 years; 80% deceased by end of follow-up) ⁵:

Lowest tertile HGS at age 85: HR 1.35 (95% CI 1.00–1.82) for all-cause mortality
Greatest relative HGS loss over 4 years: HR 1.72 (95% CI 1.07–2.77) for mortality
Effect strengthened with increasing age — at age 89, lowest two tertiles showed HRs of 2.04 and 1.73 respectively

This study is notable for focusing on the oldest-old (85+) where most standard aging biomarkers lose predictive power; HGS remained independently prognostic in this extreme-age population.

Mendelian Randomization Status

Status: partial — GWAS instruments for HGS exist, and MR analyses have been conducted for specific cardiovascular and respiratory endpoints, but a dedicated MR study of HGS against all-cause mortality or longevity has not been published.

Available evidence:

Park 2022 (n=UK Biobank + FinnGen; two-sample MR): Genetically predicted poor handgrip strength associated with OR 1.128 (95% CI 1.041–1.222) for myocardial infarction / cardiovascular endpoints — suggesting a causal (not merely associative) relationship between muscle strength and cardiovascular risk ⁶. no-fulltext-access — paper is closed-access (not_oa); OR and exact endpoint designation unverified against full text.

Farmer 2019 (UK Biobank, n=~500,000; MR component): Low genetically predicted HGS associated with increased mortality risk (HR range 1.08–1.19 in observational analysis; MR signal in same direction) ⁷.

Su 2024 (bidirectional MR): Higher genetically predicted left-hand grip strength associated with reduced chronic bronchitis risk (OR 0.35, p=0.03) and asthma risk (OR 0.78, p=0.04) ⁸.

Limitations of current MR evidence: HGS is a highly polygenic trait influenced by hundreds of loci with small effects. Individual GWAS instruments have low F-statistics, and weak-instrument bias is a concern. No published MR study has used HGS GWAS instruments against all-cause mortality or lifespan endpoints specifically. needs-replication — a large-scale two-sample MR of HGS instruments against longevity/mortality GWASs (e.g., CHARGE consortium) would resolve the causal question.

Intervention Responsiveness

Resistance training (strong positive signal)

Resistance training is the most robustly validated intervention for increasing HGS in older adults. Multiple RCTs across diverse older-adult populations demonstrate:

8–16 weeks of progressive resistance training increases HGS by 2–5 kg on average in adults aged 60+ needs-replication — specific RCT citation needed for these numeric estimates; cite from literature on verification
Effects are largest in individuals with lowest baseline strength (sarcopenia population)
Mixed (combined aerobic + resistance) training also increases HGS but effects are smaller than pure resistance training

The EWGSOP2 identifies resistance training as the primary intervention for both preventing and treating sarcopenia ².

Protein intake and dietary adequacy

Adequate dietary protein (≥1.0–1.2 g/kg body weight/day) supports HGS maintenance; protein supplementation alone in older adults with adequate baseline intake shows modest incremental benefit needs-replication. See protein-intake for the full dose-recommendation evidence base.
Leucine-rich proteins (whey, egg) may have superior anabolic signaling via mTORC1 compared to plant proteins; see mtor and deregulated-nutrient-sensing

Creatine supplementation

Creatine monohydrate supplementation (~3–5 g/day) combined with resistance training shows consistent incremental HGS improvement in older adults over training alone in several RCTs. See creatine for compound-level evidence. needs-replication — effect sizes vary widely; a systematic review/meta-analysis citation should be added on verification pass.

Physical inactivity and disuse atrophy

Bed rest, hospitalization, or disuse rapidly reduces HGS — 10 days of bed rest can reduce HGS by 2–4 kg in healthy older adults, with disproportionately slow recovery in the aged needs-replication
Post-hospitalization HGS loss is an independent predictor of prolonged functional decline

Confounders and modifiers

Factor	Direction of effect on HGS	Notes
Chronic disease burden	Decrease	Cancer, CHF, COPD, CKD all lower HGS
Acute illness / hospitalization	Decrease	Rapid disuse + inflammation
Corticosteroid exposure	Decrease	Myopathy; steroid-induced muscle wasting
Obesity	Variable	Fat mass dilutes relative strength; BMI adjustment sometimes used
Depression	Decrease	Motor slowing; motivational factors affect volitional effort
Arthritis / hand deformity	Decrease	Cannot interpret as pure muscle biomarker when hand mechanics are compromised
Dominant hand vs. non-dominant	Increase (dominant)	~10% stronger on average; protocol consistency essential

Hallmark Linkages

HGS is an integrative downstream readout of multiple upstream hallmarks operating on skeletal muscle:

stem-cell-exhaustion — Satellite cell (muscle stem cell) depletion with age impairs fiber repair and regeneration, reducing myofiber cross-sectional area and therefore maximal force output
mitochondrial-dysfunction — Mitochondrial decline reduces ATP production per muscle fiber, lowering force output per contractile unit and contributing to fatigue-based functional limitations
disabled-macroautophagy — Autophagy-dependent proteostasis maintains sarcomeric protein quality; impaired autophagy leads to accumulation of damaged myosin and actin, reducing contractile efficiency. See autophagy for mechanism.
chronic-inflammation — Elevated IL-6, TNF-α, and IL-1β suppress muscle protein synthesis (mTORC1 inhibition via NF-κB), promote protein catabolism (ubiquitin-proteasome activation), and suppress satellite cell function — collectively driving sarcopenia
deregulated-nutrient-sensing — mTORC1 activity declines in aged muscle, impairing anabolic response to amino acids; AMPK hyperactivation in aged muscle can suppress mTORC1-dependent protein synthesis

Because HGS reflects the aggregate functional output of skeletal muscle, it is simultaneously influenced by all of these hallmarks — giving it high predictive validity but low mechanistic specificity for any individual hallmark.

HGS in the Sarcopenia Phenotype Context

Cross-reference: sarcopenia — canonical molecular and physiological page.

Key relationships:

HGS is the primary functional criterion for sarcopenia (EWGSOP2); low muscle mass alone (without low strength) is now termed “pre-sarcopenia” and treated as insufficient for diagnosis
HGS correlates with whole-body muscle strength (leg press, knee extension) reasonably well (r ≈ 0.5–0.7 in community-dwelling older adults) — it is not just an upper-limb measure but a systemic muscle quality proxy
Sarcopenic obesity (low HGS with high fat mass) carries higher mortality risk than either condition alone ⁷
Low HGS is one of the five Fried Frailty Phenotype criteria; overlap with frailty diagnosis is substantial (see frailty-index)

Normative Reference Ranges

The NHANES (US) and Massy-Westropp (Australian) datasets are among the most cited population-level norms. Bohannon 2019 compared NHANES 2011–2014 (n=13,918, ages 6–80) against NIH Toolbox (n=3,594, ages 6–80) and confirmed meaningful differences by dominant/non-dominant side, sex, and age group ⁹. Importantly, the paper explicitly concludes it cannot recommend NHANES values for broad application as reference norms because the NHANES protocol (Takei dynamometer, standing, 3 trials) differs substantially from current standards (ASHT/Roberts 2011: JAMAR, seated, elbow 90°) — users should apply caution when comparing NHANES-derived tables to clinically measured values. EWGSOP2 anchors its cut-offs to UK normative data (Dodds 2014; PLoS One).

Population-level caveats:

Asian populations (AWGS consensus) use lower cutoffs: men <28 kg, women <18 kg (2019 AWGS update)
US NHANES norms tend slightly higher than European norms in the 50–70 age range
Body-size adjustment (HGS/BMI or HGS/height²) improves predictive validity for outcomes in some cohorts but is not universally required — raw values are used in EWGSOP2 cutoffs

Use as a Quarterly Tracking Metric

HGS is well-suited to personal longitudinal tracking:

Equipment: JAMAR hydraulic dynamometer ( ~~$250-350 c l ini c a l;$ 60–80 consumer alternatives available, though precision varies) or Smedley (~~$30–50)
Frequency: Quarterly tracking for aging monitoring; before/after intervention assessments for responsiveness to exercise or nutrition programs
Within-person variability: ~5% coefficient of variation; dominant hand preferred for consistency; best to test at same time of day and under similar conditions (fed, rested)
Clinically meaningful change: ~2–3 kg change is detectable beyond measurement noise in most protocols; a decline of >5 kg over 12 months warrants clinical attention

EWGSOP2 threshold proximity monitoring: Individuals approaching the diagnostic thresholds (men: 27–32 kg; women: 16–20 kg) benefit from closer monitoring and proactive intervention (resistance training, protein adequacy review).

Limitations and Gaps

Mechanistic opacity — HGS reflects the net output of multiple parallel hallmarks; a declining value does not distinguish which upstream driver is responsible. It is a functional read-out, not a mechanistic probe. no-mechanism
Hand pathology confounding — Arthritis, tendon injury, carpal tunnel syndrome, and peripheral neuropathy reduce HGS independently of muscle biology. Cannot be used as an aging biomarker when hand mechanics are compromised; an alternative (chair-stand test, knee extension dynamometry) should be substituted.
Effort dependence — HGS requires maximal voluntary effort; pain, cognition, and motivation affect results. Volitional effort variability is larger in demented or hospitalized populations.
Cutoff population specificity — EWGSOP2 cutoffs (men <27 kg, women <16 kg) are derived from European populations. Asian populations use different cutoffs (AWGS 2019: men <28 kg, women <18 kg — notably, AWGS uses a higher cutoff for men than EWGSOP2 despite generally lower Asian HGS norms). Applying one population’s cutoffs to another creates misclassification. needs-replication
Mendelian randomization gap — No published MR study has used HGS GWAS instruments against all-cause mortality or longevity phenotypes specifically. The causal path from muscle weakness to mortality (vs. weakness as a marker of disease burden) is not definitively established by genetic causal inference. needs-replication
Intervention trial endpoint underuse — Most pharmacological aging trials (rapamycin, senolytics, metformin) do not use HGS as a primary endpoint, limiting the dataset for drug-responsive signal. Growing use in sarcopenia trials (e.g., myostatin inhibitors, testosterone) but not yet in general geroprotector trials. needs-replication
Normative reference standardization — No globally accepted normative reference dataset. Massy-Westropp (Australian), NHANES (US), Dodds (UK), and various European cohort norms produce modestly different centile distributions. needs-replication

Cross-references

sarcopenia — canonical sarcopenia page; underlying biology, epidemiology, treatment
frailty-index — Rockwood deficit-accumulation index; HGS is one component (Fried overlap)
stem-cell-exhaustion — satellite cell depletion; primary upstream driver
mitochondrial-dysfunction — mitochondrial decline in muscle fiber; ATP output reduction
disabled-macroautophagy — autophagy-dependent myofibrillar proteostasis
chronic-inflammation — IL-6/TNF-α suppress muscle protein synthesis and satellite cell function
deregulated-nutrient-sensing — mTORC1 anabolic blunting in aged muscle
autophagy — mechanism by which autophagy maintains sarcomeric protein quality
mtor — mTORC1 as the master regulator of muscle protein synthesis
creatine — creatine supplementation as HGS-responsive intervention
exercise — resistance training; primary evidence-based HGS intervention

Footnotes

doi:10.1016/S0140-6736(14)62000-6 · Leong DP, Teo KK, Rangarajan S et al. (PURE investigators) · Lancet 386(9990):266–273, 2015 · observational (prospective, multinational) · n=139,691; 17 countries; median 4.0-yr follow-up · all-cause mortality HR 1.16 (95% CI 1.13–1.20) per 5-kg lower HGS; CV mortality HR 1.17 (1.11–1.24); HGS stronger predictor than systolic BP · model: community-dwelling adults 35–70 yrs, 4 continents · archive: repository HTML page obtained (not full results tables); key HRs cross-confirmed via Celis-Morales 2018 citations of PURE findings (PMID 25982160) · 100th citation percentile (1,769 citations) ↩ ↩²
doi:10.1093/ageing/afy169 · Cruz-Jentoft AJ, Bahat G, Bauer J et al. (EWGSOP2) · Age and Ageing 48(1):16–31, 2019 · systematic-review + consensus guideline · EWGSOP2 consensus: HGS cutoffs for low muscle strength — men <27 kg, women <16 kg (Table 3, p.24, confirmed against local PDF); F-A-C-S diagnostic algorithm; Jamar dynamometer recommended; HGS peaks ~age 40; strength loss 1.5–5%/yr after 50 · model: expert consensus based on European population normative data · archive: local PDF available at · 100th citation percentile (12,879 citations) ↩ ↩² ↩³ ↩⁴ ↩⁵
doi:10.1371/journal.pone.0163183 · Granic A, Davies K, Jagger C, Kirkwood TBL, Syddall HE, Sayer AA · PLoS ONE 11(9):e0163183, 2016 · observational (prospective longitudinal — Newcastle 85+ study) · n=845 (319 men, 526 women); 4 waves over 5 yr · linear annual HGS decline in men: −1.13 (SE 0.8) kg/yr (time-only model); women: curvilinear, accelerating by −0.06 (SE 0.02) kg per follow-up year above initial loss; 5-yr mean absolute change: men −5.27 (SD 4.90) kg, women −3.14 (SD 3.41) kg; high physical activity associated with 0.95 kg/yr slower decline in men · dynamometer: Takei A5401 digital (standing position) · model: community-dwelling adults aged ~85, northeast England · archive: local PDF available (gold OA PMID 27637107) ↩
doi:10.1136/bmj.k1651 · Celis-Morales CA, Welsh P, Lyall DM et al. · BMJ 361:k1651, 2018 · observational (prospective cohort) · n=502,293 UK Biobank participants; ages 40–69 yr; mean follow-up 7.1 yr (range 5.3–9.9 yr) · all-cause mortality HR per 5-kg lower HGS: 1.20 (95% CI 1.17–1.23) women; 1.16 (1.15–1.17) men; CV mortality HR: 1.19 (1.13–1.25) women; 1.22 (1.18–1.26) men (fully adjusted model, excluding events in first 2 yr) · HRs grip vs. systolic BP: grip showed stronger association with all-cause and CV mortality than systolic BP per SD · muscle weakness cutoff: <26 kg men, <16 kg women (FNIH definition) · model: UK Biobank community adults · archive: local PDF available (open access PMID 29739772) ↩
doi:10.1503/cmaj.091278 · Ling CHY, Taekema D, de Craen AJM, Gussekloo J, Westendorp RGJ, Maier AB · CMAJ 182(5):429–435, 2010 · observational (prospective cohort — Leiden 85-plus study) · n=555; age 85 at baseline; 65% women; follow-up range 8.5–10.5 yr; 444/555 (80%) deceased by end · lowest tertile HGS at age 85: HR 1.35 (95% CI 1.00–1.82, p=0.047) all-cause mortality; at age 89: HR 2.04 (1.24–3.35, p=0.005) and HR 1.73 (1.11–2.70, p=0.016) for lowest two tertiles; greatest relative 4-yr HGS loss: HR 1.72 (1.07–2.77, p=0.026) · dynamometer: JAMAR (dominant hand, standing, arm parallel to body) · model: community-dwelling oldest-old (all 85-yr-old inhabitants of Leiden, Netherlands) · archive: local PDF available (PMID 20142372) ↩
doi:10.1016/j.amjcard.2021.08.061 · Park S, Lee S, Kim Y et al. · Am J Cardiol 162:56–64, 2022 · Mendelian randomization (two-sample) · genetically predicted poor HGS: OR 1.128 (95% CI 1.041–1.222) for myocardial infarction/cardiovascular fatality endpoints (paper title: “Relation of Poor Handgrip Strength or Slow Walking Pace to Risk of Myocardial Infarction and Fatality”) · model: UK Biobank + FinnGen summary statistics · archive: not_oa (PMID 34903347) — OR and exact endpoint not verified against full text no-fulltext-access ↩
doi:10.1161/JAHA.118.011638 · Farmer RE, Mathur R, Schmidt AF et al. · J Am Heart Assoc 8(13):e011638, 2019 · observational cohort + Mendelian randomization · n=UK Biobank (~500,000) · MR: low genetically predicted HGS associated with mortality HR range 1.08–1.19; sarcopenic obesity carries higher mortality than either condition alone · model: UK adults · archive: download pending (gold OA); DOI confirmed (PMID 31221000) ↩ ↩²
doi:10.1007/s40520-024-02855-y · Su Y, Zhang Y, Zhang D, Xu J · Aging Clin Exp Res 36(1):196, 2024 · Mendelian randomization (bidirectional, univariate + multivariate) · higher left-hand HGS: reduced chronic bronchitis risk OR 0.35 (p=0.03), reduced asthma risk OR 0.78 (p=0.04) · model: MR using UK Biobank HGS GWAS instruments · archive: not checked (PMID 39395132) ↩
doi:10.5014/ajot.2019.029538 · Bohannon RW, Wang YC, Yen SC, Grogan KA · Am J Occup Ther 73(2):7302205080p1–7302205080p7, 2019 · observational (cross-sectional normative; secondary data analysis) · n=13,918 (NHANES 2011–2014, ages 6–80) + 3,594 (NIH Toolbox, ages 6–80) · confirms normative differences by dominant/non-dominant side, sex, and age group; caution: paper concludes NHANES values “cannot recommend for broad application as reference norms” due to protocol differences from ASHT/Roberts 2011 standards · model: US community-dwelling noninstitutionalized population · archive: PMC6436115 (PMID 30915969) ↩

🧬 Aging Wiki

Explorer

grip-strength-biomarker