Three Checks, One Number

Stand up right now and lift one foot off the floor.

Not a dramatic yoga pose. Just stand on one leg, arms wherever they fall, and count. Most people reading this will get past ten seconds without much trouble. Some won't. And the difference between those two groups turns out to matter more than almost anything you could learn from a mirror.

That's the whole idea behind the three tests in this week's video: a dead hang, a one-legged stand, and getting up off the floor. Five minutes, no equipment beyond something to hang from. The video showed you how to run them. This shows you what the research actually says underneath them, including the two places where the popular version of that research says more than the studies do.

I would rather tighten a claim here than let you repeat something that doesn't hold.

You have probably seen the headline: grip strength beats blood pressure at predicting death. It comes from a genuinely large study, and the headline is half right.

The PURE study followed 139,691 adults aged 35 to 70 across 17 countries, from low, middle and high income settings, for a median of four years. Every five kilograms of lost grip strength came with a 16% higher risk of dying from any cause (hazard ratio 1.16, 95% CI 1.13 to 1.20). The authors state plainly that grip was a stronger predictor of all-cause and cardiovascular mortality than systolic blood pressure (Leong et al., 2015, The Lancet).

Here is the part that gets dropped. Look at the paper's own comparison table and the picture is more mixed. For death from any cause, grip clearly outperforms blood pressure. For cardiovascular death specifically, the two are effectively tied, with confidence intervals that almost completely overlap. And for whether you go on to develop cardiovascular disease in the first place, blood pressure is the better predictor, and the paper says so.

So the honest sentence is narrower than the meme. Grip strength predicts dying better than blood pressure does. It does not predict getting heart disease better than blood pressure does. Those are different questions and they have different answers.

I have to own one here. The original video for this topic said grip beat blood pressure at predicting cardiovascular mortality specifically, and that is the one comparison where the two are level. I have corrected the video description. The claim was too strong by one word, and it was my word.

The signal itself replicates. A separate analysis of 502,293 UK Biobank participants found the same direction and a similar size, a 16 to 20% higher risk of death per five kilograms lost depending on sex (Celis-Morales et al., 2018, BMJ). That study also found something worth keeping in view: adding grip strength on top of the standard risk factors improved prediction by a very small amount. Grip is a real signal. It is not a replacement for the things your doctor already measures.

One more piece of honesty, and it is about my test rather than their research. Every study above measured grip with a calibrated dynamometer, in kilograms of force. A dead hang is not that. It is a bodyweight-dependent endurance task, and your shoulders and lats give out alongside your hands. I use the hang because you can do it in a doorway instead of a lab. Treat it as a practical stand-in for a capacity these studies measured properly, not as the tested thing itself.

The balance test has its own paper, and it is a good one. Researchers asked 1,702 people aged 51 to 75 to stand on one leg for ten seconds. Over a median seven years, 4.6% of those who could do it died, against 17.5% of those who could not. After adjusting for age, sex, body mass index and existing conditions, failing the test came with a hazard ratio of 1.84 for death from any cause (Araujo et al., 2022, British Journal of Sports Medicine).

Two things the coverage of that study usually leaves out.

The raw gap, roughly four times the death rate, is mostly age and illness. The adjusted figure, 1.84, is the one that survives accounting for those. That is still a substantial number. It is not the number in the headlines.

And this was not a population sample. Everyone in it walked into one private exercise clinic in Rio de Janeiro of their own accord. The authors describe the group as mostly white and from higher socioeconomic strata in Brazil, and they say directly that recent falls history and physical activity were never recorded. They flag both as limitations. So although the balance test feels like it should be about falling, that study did not measure falls at all. It measured death. If you have seen this test sold as a fall-risk screen, that framing did not come from this paper.

The sitting-rising test has the longest research trail of the three, and the video's version is deliberately stricter than the research version.

The original work followed 2,002 adults aged 51 to 80 for a median 6.3 years (Brito et al., 2014, European Journal of Preventive Cardiology). A newer study followed 4,282 adults aged 46 to 75 for a median 12.3 years and found the lowest scorers were 3.84 times more likely to die of natural causes and 6.05 times more likely to die of cardiovascular disease than the highest scorers, after adjusting for age, sex, body mass index and health history (Araujo et al., 2025, European Journal of Preventive Cardiology).

Both of those used a seated, cross-legged start. The version in the video starts flat on your back. Mine is the harder one. That means you should read it as a stricter bar rather than as a proven equivalent, and it means a poor score on my version is not the same event as a poor score in those studies.

I am telling you this because a test you trust for the wrong reason is worse than no test.

Mortality studies get the attention because the outcome is dramatic. The more useful research asks a quieter question: who stops being able to do their own life?

In 1995, researchers ran a three-part battery on 1,122 adults over 70 who had no disability at the start: a standing balance hold, a short timed walk, and rising from a chair five times. Four years later, the lowest scorers were 4.2 to 4.9 times more likely to have developed disability than the highest scorers, after adjusting for age, sex and chronic disease (Guralnik et al., 1995, New England Journal of Medicine).

Read that battery again. A balance hold and getting up off a seat. Two of its three parts are two of my three tests, and the outcome it measured was not dying. It was losing the ability to dress yourself, climb your stairs, carry your own shopping.

That is the actual stake. Note that the 4.2 to 4.9 figure belongs to the combined score, not to any single test on its own, and that poor performers declining is not the same claim as practising the test preventing decline. Nobody has shown the second thing.

The clinical world reaches the same place from another direction. The European consensus definition of sarcopenia, the age-related loss of muscle, makes low muscle strength its primary criterion, on the reasoning that strength is currently the most reliable measure of muscle function. Its listed consequences are the ordinary ones: falls, fractures, losing the ability to perform daily activities, loss of independence (Cruz-Jentoft et al., 2019, Age and Ageing).

You take the lowest one and you work on that.

Not all three. The instinct after a test like this is to overhaul everything, and that is the version that lasts nine days. One capacity, the weakest, trained in a way that survives a tired week.

That is the mechanism. Three cheap checks standing in for capacities that real research links to how long you live and, more usefully, to how long you keep running your own life. One of those links is smaller than the popular version. One was studied in a narrower room than you would guess. One is easier in the lab than in my video. All three are still worth five minutes of your Saturday.

The video showed you how to run the tests. This showed you what is holding them up, and where it is thinner than advertised.

If you want to score yourself properly, the full instrument, with the scoring bands and the worked interpretation of what a weak result in each capacity actually means for training, lives in Publication 02: Three Checks, One Number. This write-up was the evidence. That is the tool.