Grip strength predicts lifespan, but almost nobody measures it the same way

Handgrip strength is one of the most studied predictors of mortality risk, a quick squeeze of a dynamometer that correlates with how long people live. Huge cohort studies rely on it, and readers increasingly meet grip thresholds in health content and fitness tools. But a predictor is only as good as its measurement, and if every study measures grip differently, the risk estimates built on those numbers may not line up with each other. This review set out to catalogue exactly how mortality studies measure grip and whether the differences matter.

The authors systematically searched PubMed/MEDLINE, Web of Science and Scopus from inception to February 2022 for prospective cohort studies that measured handgrip strength objectively and reported at least one all-cause, cancer or cardiovascular mortality outcome. Forty-eight studies qualified, covering 3,135,473 participants, of whom 49.6% were women. Study quality was assessed with a standard checklist, and a regression analysis quantified how much the grip values reported by each study were influenced by which elements of the measurement protocol were controlled.

Protocols turned out to be incomplete and highly inconsistent. Half of the studies controlled body position, 39.6% controlled arm position, 33.3% elbow position, 12.5% wrist position, 13% the duration of the squeeze, 23% how the hand was adjusted to the device, and just 12.5% provided verbal encouragement. The number of attempts, which hand was tested and how the final value was calculated all varied considerably. Grip values were generally higher when fewer protocol items were controlled, higher when the number of measurements or arm position went uncontrolled, and lower when elbow position was uncontrolled or encouragement was absent.

Nothing here undermines grip strength as a predictor; the association with mortality remains one of the most consistent in the field. What this review adds is a warning about comparability: two studies, or two people, can produce different grip numbers purely because the test was run differently. The authors call for standardised protocols so that the mortality risk estimates attached to grip values become more accurate. Until that happens, cut-off values and norms drawn from different studies should be compared with some care.

If you have ever squeezed a dynamometer and checked your result against a reference table, this is a reason to treat the comparison loosely: the table's numbers came from a protocol that may not match how you tested. It is also a reason to keep your own method consistent if you track grip over time, since changes in posture, elbow angle or enthusiasm from a tester can move the reading. The underlying message that stronger grip tracks with lower mortality risk stands untouched.