Grip Strength Calculator — Percentile and T-Score by Age

Why Grip Strength Gets Measured At All

Grip strength is a 10-second test that predicts things it has no business predicting. It is associated with all-cause mortality, cardiovascular events, length of hospital stay, recovery after surgery, and the ability to keep living independently — and it does so after adjusting for age, sex and body size.

It is not that your hands matter. Grip is a cheap proxy for whole-body muscle function, and whole-body muscle function is what the outcomes track. That is why it is a core component of both the sarcopenia and the frailty phenotypes, and why a dynamometer sits in a lot of geriatric clinics.

The Two Numbers, and Why They Disagree

This page gives you a percentile and a T-score. They answer different questions and they will often point different ways.

The percentile is where you sit among people your own age and sex. The reference moves with you: an unchanged grip rises through the percentiles every year, because the comparison group is weakening. A 70-year-old man at 39 kg is at the 50th centile — the same reading at 35 would put him near the 20th.

The T-score is how far you are below the peak of young adult life, on a reference that never moves:

T = (your grip − peak mean) / peak SD

For men that peak is 51.9 kg with an SD of 9.9, for women 31.4 kg with an SD of 6.1, both at age 32. It is the same construction as the bone-density T-score used to diagnose osteoporosis, and for the same reason: what matters clinically is absolute capacity, not rank among your similarly-declining peers.

Being at the 60th centile for 80-year-olds and two and a half standard deviations below peak are both true at once. Only the second one has a cut-off attached.

Where the Cut-Offs Come From

The sarcopenia thresholds everyone quotes are not independent findings. They are this dataset:

Run the arithmetic on the peak values: 51.9 − 2.5 × 9.9 = 27.2, and 31.4 − 2.5 × 6.1 = 16.2. The European cut-offs are the T ≤ −2.5 line from this paper, rounded. They were not derived separately and then found to agree.

Two details that matter and usually get flattened:

The Asian male threshold is stricter than the European one. AWGS raised it from 26 to 28 kg in its 2019 revision while leaving women at 18. So a 27.5 kg reading is low strength under AWGS and not under EWGSOP2 — for the same man, on the same day, on the same dynamometer.

The boundary wording differs by source. EWGSOP2 and AWGS define low strength as strictly *below* their figure. Dodds describes the T ≤ −2 line as "32 kg in males and 19 kg in females, or weaker", which includes the boundary value. This calculator uses each source's own operator, which is why 32.0 kg trips one line and 27.0 kg trips two.

The Protocol Is Part of the Measurement

The reference data was built from the maximum of all trials, both hands, mostly seated. If you take the mean of three squeezes instead of the best one, you will score lower than the tables expect and read a percentile that is not yours.

Four things that move a reading more than most people assume:

• Hand grip span. A Jamar-style dynamometer has five settings. Second position is standard for most adults; the wrong setting can cost several kilograms.
• Elbow angle. Seated with the elbow at 90° is the protocol. Standing with the arm extended reads differently, and the paper's own sensitivity analysis found standing and seated protocols produced acceptably similar norms — but mixing them within your own tracking will add noise you then misread as change.
• Which hand. Dominant is typically 5–10% stronger. The norms used both hands and took the maximum, so use your best reading, not your writing hand specifically.
• The dynamometer. Different devices read differently. The paper tested this and found the normative values robust across device types — but your *own* repeat measurements should stay on one device, because the differences between models are larger than the changes you are trying to detect.

What the Reference Data Is

Dodds and colleagues, PLoS ONE 2014: 60,803 observations from 49,964 participants, 26,687 of them female, pooled from twelve general-population studies across Great Britain — ALSPAC, ELSA, the Hertfordshire studies, the Lothian Birth Cohorts, the Newcastle 85+ Study, NSHD, Understanding Society and others. Centile curves from age 4 to 90.

That makes it one of the few grip datasets that covers the whole life course rather than just old age, and it is the one the European consensus cites. It is also, usefully for anyone reading this in Britain, a British sample.

Three limits the authors state themselves, worth repeating:

1. It is cross-sectional. It compares different people at different ages, so it probably understates how fast an individual declines. The authors explicitly say the centiles should not be used to monitor an individual's trajectory. 2. Height is not adjusted for. Taller people grip harder, and these centiles do not account for it. A short person will score lower than their muscle function deserves, and a tall one higher. UK Biobank publishes height-stratified norms if that matters to you. 3. Birth years are narrow within each age band, so some of what looks like an age effect may be a cohort effect.

How This Page Computes a Percentile

The paper publishes centiles at five-year intervals. For an age in between, this calculator interpolates the published mean and SD and uses a normal distribution — which is legitimate here because the authors tested for it: "We found no evidence of skewness or kurtosis in grip strength at any age."

Checked against the paper's own 180 published centile values, that approximation sits a mean of 0.48 kg away, with 160 of 180 cells inside 1 kg and only two beyond 2 kg — both at age 20, the sparsest adult band in the dataset. The published row for the nearest age is printed beside the interpolated ladder on this page so you can see the source numbers rather than only the model.

If the Reading Is Low

A low grip is a reason to ask a question, not an answer. Sarcopenia is not diagnosed on grip alone — every published criterion pairs low strength with either low muscle mass or low physical performance, and AWGS's "possible sarcopenia" category exists precisely to mark a screening result as not yet a diagnosis.

What it is good for: a baseline. A single reading tells you where you sit. The same reading taken annually, same device, same protocol, tells you something no percentile can.