Hand grip strength as the all-cause-mortality predictor — what the data really says

The Leong and PURE 2015 finding

Leong and colleagues published the headline analysis from the Prospective Urban Rural Epidemiology study in The Lancet in July 2015. They followed 142,861 adults aged 35 to 70 across 17 countries — low-income, middle-income, and high-income — for a median of four years. The variable of interest was hand grip strength, measured with a calibrated handheld dynamometer. The outcomes were cardiovascular disease, non-cardiovascular disease, and all-cause mortality.

The result was striking: each five-kilogram reduction in grip strength was associated with a 17 percent increase in risk of cardiovascular mortality, a 17 percent increase in risk of all-cause mortality, a 9 percent increase in stroke risk, and a 7 percent increase in heart-attack risk. The relationship held across all 17 countries, across income levels, across both sexes, and after adjustment for age, smoking, education, employment, physical activity, and other usual confounders. Grip strength was a more powerful predictor of cardiovascular death than systolic blood pressure in the PURE cohort.

That last sentence is the one that ran in the headlines. It is also the one that got most often misinterpreted. Grip strength does not cause cardiovascular death. It correlates with it. The mechanism is downstream, not direct.

Why grip predicts what it predicts

A handheld dynamometer measures the maximal force of the hand and forearm muscles. The hand and forearm are not unusually important to the cardiovascular system. So why does grip predict so much?

The answer has two parts. First, grip strength is a proxy for total-body muscular strength. The same biological and behavioural factors that build a strong grip — protein intake, sleep, physical activity, neuromuscular health, hormonal status — build strong legs and a strong trunk. Measure grip and you measure, indirectly, the muscular reserve of the whole body. Sarcopenia — age-related muscle loss — shows up at the hand before it shows up at the leg in some cases, and consistently in older adults.

Second, low grip strength reflects integrated frailty in ways that single organ-system measures do not. A person whose grip is in the bottom quartile for their age and sex is, on average, also lower on cardiovascular fitness, balance, bone density, sleep quality, and protein nutrition. They are not lower on all of them; they are lower on enough of them that the composite signal is strong. Grip is a clinical-friendly thermometer reading of overall biological resilience.

Sarcopenia and grip — Cruz-Jentoft 2019

Cruz-Jentoft and colleagues published the revised European consensus definition of sarcopenia (EWGSOP2) in Age and Ageing in 2019. They moved grip strength from a confirmatory measure to a primary diagnostic criterion. In the new framework, low muscle strength — operationalized as low grip — is sufficient to begin a sarcopenia workup. Low muscle quantity or quality confirms the diagnosis; low physical performance establishes severity.

This was not an arbitrary choice. The 2019 consensus reviewed roughly two decades of research on the predictive validity of different sarcopenia measures and concluded that grip strength was the single best one-shot screen. Cheap, repeatable, equipment-light, internationally standardized. The change in 2019 meant that, in clinical practice across most of Europe and increasingly in Canada, a low grip score in an adult over 65 now starts a workup for muscle loss. That is a meaningful elevation of grip from "interesting biomarker" to "diagnostic anchor."

The 26-kg and 16-kg cut-points

The numbers that matter for adults are the EWGSOP2 cut-points: less than 27 kg of maximal grip in men, and less than 16 kg in women, indicates probable sarcopenia. These thresholds are for the dominant hand and are intended for adults over 65, but the gradient is continuous — adults under 65 with similar grip values are also at meaningfully elevated risk.

For context, a healthy untrained adult man in his 40s typically grips around 45-55 kg on the dominant hand. A healthy untrained adult woman in her 40s typically grips around 27-35 kg. Athletes and manual workers grip considerably higher; rock climbers can exceed 60 kg without specifically training grip. The 27/16 cutpoints represent the lower edge of the population distribution, where mortality and disability risk inflect upward.

Dynamometers are inexpensive and widely available. A single measurement does not establish a trend, but a measurement once a year — taken under the same conditions, on the same hand, with the same technique — gives a real signal about whether your muscular reserve is holding, building, or eroding.

Training grip after 50

Grip strength responds to training at any age. The 2019 sarcopenia consensus and subsequent intervention studies (notably Liu and Latham 2009, Cochrane Review on resistance training in older adults) have shown that grip-specific work — sustained holds, farmer's carries, pulling movements — produces measurable improvements in adults into their 80s and 90s. The trainability is remarkable. The application is straightforward.

Effective grip training does not require specialized equipment. Three exercises cover most of the adaptation: farmer's carries (walk while holding heavy weights at the sides, 30-60 seconds per set, 3 sets, 2-3 times per week), dead hangs from a pull-up bar (start with a few seconds, build to 30+ seconds, 3 sets), and slow-tempo gripping work with a hand-grip trainer or thick towel-wringing motion. Two to three sessions per week, 10-15 minutes per session, builds noticeable strength within 8 weeks.

The mistake older adults make is treating grip as a fragile thing to be protected. The data point the other way: grip is robust, trainable, and one of the most valuable single capacities to maintain through aging.

Grip and cognitive decline (Sayer 2015)

Sayer and Kirkwood published a related-but-distinct line of work on grip strength and cognitive trajectories. Their 2015 BMJ paper synthesized evidence from longitudinal studies showing that lower mid-life grip strength was associated with steeper cognitive decline in later life. The association persisted after adjustment for education, socioeconomic status, and baseline cognition.

The mechanism is not that gripping a dynamometer trains the brain. The mechanism is shared upstream factors — vascular health, inflammation, neuromuscular integrity, physical activity history — that affect both muscle and brain. Grip is once again a proxy, this time for the kind of integrated biological aging that affects both grey matter and skeletal muscle. The clinical implication: an adult in their 50s with a deteriorating grip score has a signal worth taking seriously, not because the hand matters but because the hand is reporting on the whole.

Everyday grip activities that count

Most of the grip work an adult body needs can be embedded in normal life rather than added as a separate training session. Carrying groceries from the car in one trip rather than two. Carrying a heavy bag through an airport rather than rolling it. Gardening — pulling weeds, turning soil, carrying compost. Kayaking, canoeing, paddleboarding (the paddle is gripped continuously). Opening jars. Wringing out a wet towel. Climbing trees with kids.

The point is not that these replace structured strength training. The point is that grip is exposed to use in normal life in a way that other muscle groups are not, and choosing the higher-grip option in routine activities meaningfully maintains the capacity. A cottage week with kayaking, dock-work, and carrying firewood is a grip-training week, whether or not it is labelled that way.

When weak grip is a medical signal

A grip score below the EWGSOP2 cutoff in an adult over 65 is a signal worth discussing with a family physician. So is a grip score that has dropped by more than 10 percent in a year in an adult of any age. So is a meaningful asymmetry between hands — more than 10 percent difference, in the absence of an obvious injury or dominant-hand explanation — which can be an early sign of neurological problems.

This is not catastrophizing. It is treating grip as the clinical biomarker the literature has shown it to be. Most low grip scores have ordinary, treatable explanations: insufficient protein intake, low physical activity, poor sleep, undertreated depression, recovering from illness. Some have less ordinary explanations: early Parkinson's signs, peripheral neuropathy, rheumatologic conditions. The grip test does not diagnose; it flags. The conversation with the physician is where the diagnosis, if there is one, happens.

Practical takeaways

• PURE 2015 produced the most powerful single-biomarker mortality association in the modern epidemiology literature.
• Grip predicts because it proxies total muscular strength and integrated biological resilience, not because the hand itself is cardiovascular.
• The EWGSOP2 thresholds — 27 kg for men, 16 kg for women — are clinical anchors; below these, sarcopenia workup begins.
• Grip is trainable at every age; farmer's carries, dead hangs, and tempo gripping are the most effective specific interventions.
• A grip drop of more than 10 percent in a year, or a side-to-side difference of more than 10 percent without explanation, is worth a physician conversation.

Extended takeaways

The reason grip strength has emerged as the most useful single-measure biomarker for adult health is not that the hand is special. It is that the hand is accessible. Measuring cardiovascular fitness rigorously requires a treadmill, gas analysis, and trained staff. Measuring leg strength requires equipment most clinics do not have. Measuring grip requires a $40 dynamometer, 90 seconds, and a clinician who knows how to administer the test. That accessibility is what has driven grip from an interesting research variable into a clinical anchor. The most predictive biomarker the literature has produced is also the easiest one to actually use.

What the PURE data and the subsequent literature have not done — and could not do — is establish that improving grip strength directly improves the mortality outcomes the data are correlated with. The intervention question is different from the observation question. We know that low grip predicts higher mortality. We do not yet know, in randomized-controlled-trial terms, that taking a 60-year-old with low grip and training their grip up to the population median reduces their mortality risk. The plausibility is high — grip-training programs also improve total muscle strength, balance, and metabolic markers — but the direct mortality RCT has not been run and likely never will be.

What this means practically for an adult reader is that grip is best understood as a window into the systems that actually matter. If your grip is dropping, the answer is not just to train grip — though you should. The answer is to look at the other inputs grip is reporting on: protein intake, sleep, total physical activity, sedentary time, alcohol, depression, recent illness. Improve those, and grip will improve as a downstream consequence. Improving grip in isolation, without addressing what is producing the weakness, is treating the gauge instead of the engine. The gauge is still worth reading, often.

Frequently asked questions

How do I measure my grip strength at home?

With a calibrated handheld dynamometer. Squeeze as hard as you can for 3-5 seconds, take the best of three trials, on the dominant hand. Same time of day, same hand, ideally not after exercising.

Is one hand stronger than the other always?

The dominant hand is typically 5-10 percent stronger than the non-dominant hand. Differences greater than 10 percent in the absence of an injury reason are worth investigating.

Does grip training really transfer to anything else?

It transfers most clearly to other grip-dependent tasks — carrying, climbing, opening, lifting. It does not directly build leg strength or cardiovascular fitness. The value is the grip itself and what grip indicates about whole-body resilience.

I have arthritis in my hands. Should I still train grip?

Yes, with adjustments. Therapy putty, low-resistance grippers, and isometric holds (holding an object without dynamic squeezing) are usually tolerated. A hand therapist can help calibrate. Not training grip is worse for arthritic hands than carefully training it.

How does grip strength change with age?

Peak grip arrives in the late 20s to early 30s. Decline begins in the 40s and accelerates after 60. The decline is partially modifiable through training — trained older adults often have grip strengths comparable to untrained adults 20 years younger.

References