Grip strength by age: what is normal, and which chart to trust

The most used grip chart pools 12 British studies: 60,803 measurements from 49,964 people aged 4 to 90 (Dodds 2014). It gives centiles. The 50th centile, the median, is the middle person. The 10th centile is weaker than nine in ten people of that age and sex; the 90th is stronger than nine in ten.

Selected values, in kg (10th / 50th / 90th centile):

• Men at 45: 36 / 49 / 61
• Men at 65: 31 / 43 / 53
• Men at 85: 19 / 29 / 38
• Women at 45: 22 / 30 / 38
• Women at 65: 17 / 25 / 33
• Women at 85: 11 / 17 / 23

The peak comes early. Mean grip was highest at age 32: 51.9 kg in men and 31.4 kg in women. The spread is wide at every age. A 45-year-old man at the 10th centile and one at the 90th differ by 25 kg, which is more than the median falls between 45 and 85.

Canada measured 11,108 healthy people aged 6 to 79 in its national health survey, standing, on a Smedley spring dynamometer (Wong 2016). Men’s median grip is 48.4 kg at 40–44, 45.9 kg at 55–59 and 33.7 kg at 75–79. Women’s is 28.5 kg at 40–44, 25.9 kg at 60–64 and 21.1 kg at 75–79. For women these run 2–3 kg below the British centiles.

The largest chart is international. The iGRIPS review pooled 100 studies and 2,405,863 adults in 69 countries, and adjusted every result to one protocol (Tomkinson 2025). It puts the peak at 49.7 kg for men and 29.7 kg for women, at 30–39. Men’s medians are 48.8 kg at 40–44, 42.4 kg at 60–64 and 32.3 kg at 80–84; women’s are 29.4, 26.2 and 20.4 kg.

US norms from the NIH Toolbox, in 1,232 people, give mean grip from 49.7 kg (dominant hand, men 25–29) down to 18.7 kg (non-dominant hand, women 75–79) (Wang 2018). Only the abstract could be checked, so we treat it as a cross-check, not a chart.

Which to trust? Tomkinson is the broadest and the most consistent in method, but its values are rebuilt by simulation from published summaries. Dodds is real British data, large, and robust to device type within 10%. Neither is wrong. The honest answer is that "normal grip" is a range of a few kilograms, not one number.

In a randomised cross-over study, 118 adults aged 45–74 squeezed four devices. Electronic dynamometers read 4–5 kg higher than the hydraulic Jamar or the Smedley spring device (Lessof 2023). In that study the Jamar and the Smedley differed by only 0.2 kg. Other studies found the Smedley reading 2.6–3.1 kg lower than the Jamar in people aged 69–89, and 4 kg lower in a group with a mean age of 76.

Even when the average gap is small, one person can differ by more, because the limits of agreement between devices were wide in every pair. None of these studies tested the low-cost digital dynamometers most people buy.

Protocol matters too. Dodds mostly took the best of six squeezes across both hands. Canada took the best of two per hand, standing. Tomkinson’s reference is seated, elbow bent, three squeezes per hand. Its adjustment factors between protocols ranged from under 1% to 10% for device and position, and up to 17% for how the result was reported.

The European sarcopenia definition (EWGSOP2) treats grip under 27 kg in men and under 16 kg in women as low strength (Cruz-Jentoft 2019). Those cut-offs come from Dodds, where they mark a T-score of −2.5 or lower. At 80, 23.0% of men and 26.6% of women in the British data were below them. UK Biobank used nearly the same line: under 26 kg for men and under 16 kg for women (Celis-Morales 2018).

These are diagnostic thresholds for older adults. They are not targets for someone of 40. A 45-year-old man at 30 kg is above the sarcopenia line and still below the 10th centile for his age.

Every chart above is cross-sectional. It compares different people at different ages, not the same people ageing. The Dodds authors warn that their centiles should not be used to monitor one person’s change over time, because cross-sectional data are likely to underestimate individual decline.

As a rough guide to the population slope, Tomkinson puts the typical fall per decade at 2.8 kg (men) and 1.4 kg (women) in middle adulthood, and 5.6 kg and 3.5 kg in late adulthood. Individuals do not follow it closely. In the Baltimore Longitudinal Study of Aging, over a mean of nine years, grip did not decline at all in 48% of people under 40, 29% of those aged 40–59 and 15% of those over 60 (Kallman 1990).

The grip test in the app is standing, arm by your side, elbow straight, dominant hand, best of three. That is close to the Canadian survey and differs from Dodds and Tomkinson, which took the best of both hands. Whatever protocol you use, keep it the same every quarter.

The app scores grip against Dodds, by decade of age. Because the source is continuous and the app uses ten-year bands, a percentile can step when you turn 40, 50 or 60. The app records which device you used with each result and asks you to stay on it, so your own line stays comparable even if the percentile is off by a few kilograms. Change is tracked from your own results, never from the slope of the chart.

To see where a number sits, the calculator shows the percentile and its source. It is a reference point, not a diagnosis, and none of this is medical advice.