Grip is a cheap aging marker. Farmer carries are honest work — not a longevity drug.

The fight

One cheap, repeatable number: how hard you can squeeze a calibrated dynamometer. In men, under 27 kg is European consensus for probable sarcopenia — a reason to look at whole-body strength, gait, illness, protein and training. The other story is gripper-as-personality, gadgets that only train the crush, or the leap from “weak grip predicts death” to “farmer carries extend life.” Epidemiology is not an intervention. Both of the big papers, opened, refuse causal language.

I am not in a position to diagnose sarcopenia from a blog. I can tell you what PURE and UK Biobank measured, what EWGSOP2 uses the cut-point for, and how empty the farmer-carry outcome cupboard is.

What low grip actually predicts

PURE, Leong and colleagues, Lancet 2015. Community households, at least one member aged 35–70, 17 countries. Grip: Jamar dynamometer; mean of maximum values from each hand. Analysis n=139,691 with known vital status, of 142,861 enrolled. Median follow-up 4.0 years (IQR 2.9–5.1). Deaths: 3,379 (2.4%). Adjusted Cox and competing-risk models for age, sex, education, employment, physical activity, tobacco and alcohol, daily dietary energy, percent calories from protein, self-reported hypertension, diabetes, heart failure, CAD, COPD, prior stroke or cancer, BMI, waist-to-hip ratio. Per 5 kg reduction in grip: all-cause mortality HR 1.16 (95% CI 1.13–1.20); cardiovascular mortality 1.17 (1.11–1.24); non-cardiovascular mortality 1.17 (1.12–1.21); MI 1.07 (1.02–1.11); stroke 1.09 (1.05–1.15). After adjustment, no significant association with incident diabetes (HR 1.03, 0.996–1.06), hospital admission for pneumonia or COPD, or fracture; fall injury HR 0.968 (0.939–0.998). Cancer association was heterogeneous by country income — in high-income countries, subdistribution HR 0.916 (0.880–0.953), that is higher grip with higher cancer incidence; not seen in middle- or low-income countries. Post-hoc per-standard-deviation comparison: grip predicted all-cause death more strongly than systolic blood pressure (HR 1.37 versus 1.15 per SD). That is a prediction comparison. It is not “grip is more important than blood pressure to treat.” Reverse-causation checks: excluded deaths within 6 months; excluded baseline CVD and cancer; those analyses “yielded similar findings” and “partly allay concerns over reverse causation.” Discussion, in their own words: “The observational nature of this study does not allow us to make strong conclusions on the causal role of muscular strength”; residual confounding cannot be excluded. Last line: measurement is a simple risk-stratifier; “Further research is needed to identify determinants of muscular strength and to test whether improvement in strength reduces mortality and cardiovascular disease.” That trial, with farmer carries or grippers, is not in the opened set.

UK Biobank, Celis-Morales 2018, later confirmed the signal and still would not call it causal. n=502,293 with grip (mean of right and left, kg). Mean follow-up 7.1 years (range 5.3–9.9) for mortality. Deaths 13,322 (2.7%). Fully adjusted model 3 plus model 4 landmark — exclude events in the first 2 years. Covariates include age, deprivation, ethnicity, month of recruitment, comorbidities, height, BMI categories, smoking, physical activity, sedentary time, diet. Per 5 kg lower grip, landmark hazard ratios (women, then men) included all-cause mortality 1.20 and 1.16; CVD mortality 1.19 and 1.22 (the text also reports 1.24 for men in one sentence). Muscle weakness by FNIH cut-points — ≤26.0 kg men, ≤16.0 kg women — associated with higher hazard for most outcomes except some cancer subtypes. Adding grip to an office-based score (age, sex, diabetes, BMI, SBP, smoking) improved C-index: all-cause +0.013, CVD death +0.012, incident CVD +0.009. Limitations they print: UK Biobank is not fully representative; “Reverse causality is possible in any observational study; although our results were similar after a landmark analysis… we cannot exclude the possibility of reverse causality. Similarly, residual confounding is always possible and the associations observed may not imply causality.” Because the aim is prediction, reverse causality is “not a major limitation” for that use. Prediction is not a farmer-carry prescription.

Grip also predicts future disability, still as epidemiology. Bohannon 2019, a narrative review, is the function synthesis. Grip is “largely consistent” as an explanator of concurrent overall strength, upper-limb function, BMD/fractures/falls, nutrition, cognition, depression, multimorbidity, quality of life; and as a predictor of all-cause and disease-specific mortality, future function, fractures, hospitalisation. Caveat he prints: grip “may not always be reflective of overall strength” and may be better with a lower-limb measure. Concurrent function: Wang and Chen cut-offs to manage heavy tasks (lift or carry 11 kg objects) 18.5 kg women / 28.5 kg men. Walking-limitation thresholds across 6 studies: men 23.2–39.0 kg, women 15.9–22.0 kg. His 2008 review covered 9 studies; function measured a median of 10 days to 25 years later. For the 25-year cohort — his citation 103, Rantanen JAMA 1999; the JAMA PDF did not open on this pass, so these numbers are Bohannon’s account of that paper — lowest baseline grip → higher odds of walking ≤0.4 m/s (OR 2.77), unable to rise from a chair (2.73), lifting 4.5 kg (1.94), heavy housework (1.69), dressing (2.43), bathing (2.06); not significant for walking 0.8 km, 10 steps, eating, toileting as listed. Dodds: mid-life grip → mobility and/or personal-care disability in early old age OR 1.84. McGrath: higher baseline grip, lower odds of later ADL (0.95) and IADL (0.92) disability in older Mexican Americans. Cooper 2010, opened, is a mortality meta, not disability: community-dwelling adults, weakest versus strongest quarter of grip, 14 studies, 53,476 people, summary HR 1.67 (1.45–1.93) after age, sex and body size (I² 84%). Association also present in five studies with mean age under 60. None of that is a training trial. Do not collapse “predicts disability” into “farmer carries prevent nursing homes.”

27 kg is probable sarcopenia, not a death line

EWGSOP2, Cruz-Jentoft and colleagues, 2019. Sarcopenia is a “progressive and generalized skeletal muscle disorder” associated with falls, fractures, physical disability and mortality. The 2019 European consensus moved low muscle strength to the front of the diagnosis. Algorithm F-A-C-S: Find–Assess–Confirm–Severity. Find: SARC-F. Assess: grip or chair stand. Confirm: DXA or BIA in clinic. Severity: gait, SPPB, timed up-and-go, 400 m walk. Table 3 in the opened author manuscript: grip <27 kg men / <16 kg women (Dodds 2014; T-score ≤ −2.5 versus young adults); chair stand >15 s for 5 rises; appendicular skeletal muscle <20 kg men / <15 kg women; ASM/height² <7.0 kg/m² men / <6.0 kg/m² women in this manuscript table; gait ≤0.8 m/s; SPPB ≤8; TUG ≥20 s. “In clinical practice… low muscle strength… is enough to trigger assessment of causes and start intervention.” Grip “correlates moderately with strength in other body compartments, so it serves as a reliable surrogate”; Jamar is “validated and widely used.” Time course: mass and strength peak by about 40 years, then decline; beyond 50 they cite 1–2% per year leg mass loss and 1.5–5% per year strength loss. “Interventions including nutrition and exercise training seem to slow or reverse these processes” — that citation is not unpacked here as farmer-carry data. A 2019 published correction exists; that PDF did not open. I am not treating the women’s ASM/height² cell as gospel without it. 27 kg is a rounded −2 SD convenience cut, not a metaphysical cliff. FNIH weakness in UK Biobank was ≤26 kg in men. Cut-points disagree. Confirmation of sarcopenia still needs quantity or quality. Isolated gripper work does not treat the diagnosis.

Training the marker is not a mortality trial

You can change a dynamometer score. Transfer from mixed exercise in healthy older adults is small. Labott 2019: 24 trials, PEDro mean 5.8 ± 0.9, n=3,018, mean age 73.3 ± 6.0, healthy community-dwelling adults 60 and over. Pooled standardised mean difference 0.28 (95% CI 0.13–0.44), I² 56%. Conclusion: “Meaningful but small transfer effects of a multitude of different training approaches on handgrip strength… Handgrip strength cannot clearly be recommended to assess general functional performance for all kinds of exercise programs, whereas task-specific training and multimodal training modes seem to provide an appropriate stimulus to also improve handgrip strength.” Small SMD is not “therefore buy a gripper.”

ACSM’s 2026 Position Stand (Currier and colleagues) is the parent program. Overview of 137 systematic reviews, more than 30,000 participants, RCTs, healthy adults 18 and over, resistance training at least 6 weeks versus no exercise or alternate RT. Versus control, RT improved muscle strength, size, power, endurance, contraction velocity, gait speed, balance, and multiple physical function outcomes. Strength favored heavier loads (≥80% 1RM), full range of motion, 2–3 sets, earlier in the session, at least 2 sessions a week. Hypertrophy favored at least 10 sets a week and eccentric overload. Power: moderate loads 30–70% 1RM, fast concentric. Training to failure, equipment type (machines versus free weights), exercise complexity and periodisation did not consistently impact outcomes. Biggest practical line: moving from no RT to some RT; individualise for adherence. Safety: in an analysis they cite of more than 38,000 participants (more than 6,700 RT; more than 11,000 older), exercise did not increase serious adverse events; nonserious adverse events listed as pain, fatigue, bursitis, edema. The intro also states RT is associated with reduced mortality and CVD/cancer/diabetes risk via the reviews it cites — that is resistance training generally, not carries, not handgrip as the mediator they proved. ACSM 2026 does not name farmer walks, does not use HGS cut-points as the outcome, and does not claim RT is a longevity drug because grip went up. Do not hang PURE’s hazard ratio on ACSM’s gait-speed improvement.

A loaded carry is a walking, anti-lateral-flexion, torso-stiffness task. A gripper is a hand task. The opened carry papers are electromyography and biomechanics in young people, not RCTs in men 45–65. Ellestad 2024: n=18 healthy college-aged (12 women, 6 men; mean age about 20.7–20.8 years). Surface EMG: rectus abdominis, external oblique, longissimus, multifidus, versus plank, farmer’s carry, farmer’s hold, suitcase carry, suitcase hold. Farmer’s carry / hold load 50.7 ± 1.9 kg split across two dumbbells. Walking the load beat holding it: farmer’s carry greater than hold bilaterally in longissimus, multifidus, rectus abdominis, external oblique (longissimus +9.2% left / +10.8% right versus hold). Suitcase walk greater than suitcase hold for longissimus and multifidus both sides; extra ipsilateral rectus and oblique. Plank still crushed farmer’s carry on rectus and oblique (rectus about +38–46% versus carry). Authors: farmer’s and suitcase carries are “functional,” related to carrying tasks. They do not measure grip kilograms, gait speed, sarcopenia, or 12-week training. McGill 2009, opened abstract: EMG plus spine model in strongman events including farmer’s walk, suitcase carry, keg walk, super yoke, Atlas stone, tire flip, log lift. Carrying events challenged different abilities than lifting events; “loaded carrying would enhance traditional lifting-based strength programs.” Super yoke produced the highest spinal loads. Suitcase and asymmetric carry: unique lumbo-hip demand. Technique and load mechanics in competitors, not a community RCT. I did not open an RCT of farmer carries versus control for mortality, disability, EWGSOP2 status, or even 12-week handgrip in men 45–65. That cupboard is empty. I am not going to pad it with Strongman YouTube.

Do you need to buy anything

A Jamar, or an equivalent calibrated dynamometer, if you want a number. Dumbbells, kettlebells or a trap bar if you want the carry. Whole-body resistance training at least 2 days a week, major muscle groups, is what ACSM 2026 actually supports for strength and physical function. A farmer or suitcase carry is a legitimate loaded walk that, on EMG, is not a plank and not a gripper: it asks the torso to stay tall while the legs walk. That is honest training for carrying groceries and suitcases. Scale the load so the walk stays a walk — posture, not a shuffle.

I will not sell the marker as the medicine. Beating 27 kg with a Captains of Crush is not beating aging. Harvesting PURE’s 16% per 5 kg with 8 weeks of carries is not a finding anyone opened. Gadgets that only train the crush are a personality, not a program.

When this is not the article

Under 27 kg on a calibrated dynamometer is a reason to look for causes — illness, inactivity, protein, gait — and to start intervention, not a reason to max a carry. If you have known heart disease, Williams and the AHA 2007 statement is the honest brake: Valsalva (forced expiration against a closed glottis) plus high muscle tension raises intrathoracic pressure, drops venous return, can drop cardiac output; on release, venous return and cardiac output surge into a constricted arterial tree; the blood-pressure rise is “potentially quite dramatic.” Lightheadedness during; headache after; “In patients with heart disease, symptoms of myocardial ischemia may ensue as a result of elevated BP and increased myocardial work.” MacDougall 1985, five experienced bodybuilders, intra-arterial blood pressure: lifts to failure at 80–100% max. Greatest peaks: double-leg press mean 320/250 mmHg; one subject over 480/350 mmHg. Single-arm curl to failure: group mean peak 255/190 mmHg. Mouth pressures 30–50 Torr on a max lift or near failure — part of the rise is Valsalva. That is five young bodybuilders, not a 60-year-old with uncontrolled hypertension doing suitcase carries. AHA does not ban all loaded carries; it flags strain, Valsalva and heart disease. Exhale on the effort. Do not make the carry a max-strain circus.

ACSM 2026: resistance training did not increase serious adverse events in the analyses cited; nonserious: pain, fatigue, bursitis, edema. Coronary-disease RT studies: musculoskeletal complications often from preexisting conditions (example: knee arthritis), managed by changing intensity or position. I did not open a paper that lists “farmer carry” as a named contraindication for acute tendinopathy. Reasonable clinical caution — don’t crush an angry tendon; don’t heavy-Valsalva through uncontrolled hypertension — is not a trial result. I will not write a fake orthopedic protocol.

That list is not me diagnosing you. A GP, a physiotherapist or a cardiologist will get further with an unexplained drop in strength, or with heart disease plus a closed-glottis strain, than a gripper will.