The Research
Each finding is one line. Tap any of them to see the study, who was in it and what it found, with a link to the published paper.
01 · Effort
Almost every gym decides your weight, your sets and when you stop by one instrument: how it felt. Researchers have tested that instrument for decades. Here is what they found.
Researchers secretly watched 44 adults through an hour of activity they chose themselves, then asked them to recall it. Recall was only moderately accurate. They underestimated the time they spent sitting and overestimated their aerobic activity by more than 300%. Men overestimated more than women.
Klesges et al., Medicine & Science in Sports & Exercise 1990;22(5):690–697
In the Look AHEAD trial, 3,582 adults with type 2 diabetes rated their effort at the level that is supposed to mean moderate. Their heart rates were checked at the same time. 57% were in the target range. 37% were working harder than intended and 6% were working easier.
Unick et al. (Look AHEAD), Journal of Novel Physiotherapy and Physical Rehabilitation 2014
Asking a lifter how many reps they have left is called reps in reserve, and it is the standard way gyms decide when a set is hard enough. A review pooled 12 studies and 414 people. On average they underpredicted the reps they had left by about one. Their guesses improved only as the set got close to failure, and years of training did not make anyone better at it.
25 well-trained men squatted 70% of their maximum to failure and called out how many reps they had left as they went. When they called five left, they were off by about five. Close to failure they were off by about two. The longer the set, the worse the guesses got.
Zourdos et al., Journal of Strength and Conditioning Research 2021;35(2S):S158–S165
160 trained men chose the weight they would use for a set of ten, then did as many reps as they could with it. Only 22% finished between ten and twelve. 78% got thirteen or more, and the average was sixteen. A separate review of 18 studies found that when people choose their own weights, they choose about half of their maximum.
Barbosa-Netto et al., Journal of Strength and Conditioning Research 2021;35(2S):S166–S172 · Steele et al., Sports Medicine 2022;52(12):2909–2923
54 adults over 45 kept a diary of a 12-week home strengthening program while a hidden accelerometer counted the same exercises. The diaries recorded a median of 220 exercises. The accelerometers counted 176.
Nicolson et al., Journal of Orthopaedic & Sports Physical Therapy 2018;48(12):943–950
Every one of these errors comes from the same place, a person judging their own effort.
The whole gym runs on an AI-based strength measurement system. Artificial intelligence reads exactly what level of effort you are putting in on every repetition, and it makes sure you are in the right strength-building and muscle-building zone at all times. No junk reps.
02 · The six markers
Each of the six markers in the Longevity Analysis has been followed against how long people live. Here is the research behind each one.
The PURE study followed 139,691 adults in 17 countries. Every 5 kg drop in grip strength came with a 16% higher risk of dying from any cause and 17% from heart disease. In the authors’ words, grip strength was a stronger predictor of both than systolic blood pressure. An earlier review of 14 studies found the weakest quarter of people died at a 67% higher rate than the strongest quarter. The largest review since, 48 studies and 3.1 million people, found the risk kept falling as grip went up across the whole range it studied.
Leong et al. (PURE), Lancet 2015;386:266–273 · Cooper, Kuh & Hardy, BMJ 2010;341:c4467 · López-Bueno et al., Ageing Research Reviews 2022;82:101778
Researchers measured muscle mass relative to height in 3,659 older adults and followed them for 10 to 16 years. The most muscular quarter died at a rate about 20% lower than the least muscular quarter. A second study of 11,687 adults found more limb muscle went with lower mortality, independent of body fat and other risk factors.
Srikanthan & Karlamangla, American Journal of Medicine 2014;127(6):547–553 · Abramowitz et al., PLOS ONE 2018;13(4):e0194697
A dose-response review of 72 prospective studies found every common measure of central fat, including waist size and waist-to-hip ratio, rose with the risk of dying from any cause. This is why we read body fat on an instrument.
The Copenhagen Male Study followed 2,798 men for 16 years. Risk rose 16% with every ten beats per minute, and men above 90 died at about three times the rate of men at 50 or below. The lead author told reporters that men at 80 or above were likely to die four to five years earlier than men at 65 or below. A review of 46 studies and more than 1.2 million people confirmed the pattern across both sexes.
Jensen et al., Heart 2013;99(12):882–887 · Zhang, Shen & Qi, CMAJ 2016;188(3):E53–E63
1,702 adults aged 51 to 75 were asked to stand on one leg for ten seconds and were followed for a median of seven years. One in five could not do it. 17.5% of them died, against 4.6% of the people who could. After adjusting for age, sex and health conditions, their risk was 84% higher. A Mayo Clinic study found one-leg standing time falls faster with age than grip or leg strength. We test it with your eyes closed. That removes vision and tests your vestibular and proprioceptive systems on their own.
Araujo et al., British Journal of Sports Medicine 2022;56(17):975–980 · Rezaei et al. (Mayo Clinic), PLOS ONE 2024;19(10):e0310764
A rounded upper back is called hyperkyphotic posture. Among 1,353 older adults followed for about four years, the ones with it died at a 44% higher rate after adjusting for age and sex. Our posture screen is a movement assessment that finds what your body is compensating for, and why something aches.
Kado et al., Journal of the American Geriatrics Society 2004;52(10):1662–1667
The Longevity Analysis reads all six on instruments in twenty-five minutes, turns them into one score out of 600, and runs them again every six weeks so you can watch them move. It is free.
03 · Muscle over time
The loss of muscle with age is called sarcopenia. The loss of strength with age is called dynapenia. Both start earlier than people think, and both respond to the right training.
A review of how muscle changes with age estimates a loss of 3 to 8% of muscle mass per decade after 30, with the rate climbing after 60.
Volpi, Nazemi & Fujita, Current Opinion in Clinical Nutrition and Metabolic Care 2004;7(4):405–410
The Health ABC study measured the same older adults year after year. Their strength fell several times faster than their muscle mass, so the muscle that remained produced less force. This is why we measure strength directly, on an instrument.
Goodpaster et al., Journals of Gerontology A 2006;61(10):1059–1064
Ten frail nursing-home residents with an average age of 90 trained heavy, at 80% of their maximum, for eight weeks. Their strength rose 174% on average and their thigh muscle grew about 9%. The oldest was 96.
A review of 25 controlled trials in 819 people aged 60 to 90 found the largest strength gains came from loads of 70 to 79% of maximum. A 2025 review of heavy training in older adults put the average gain at about 2.5% per session.
Borde, Hortobágyi & Granacher, Sports Medicine 2015;45(12):1693–1720 · Tøien et al., Journal of Cachexia, Sarcopenia and Muscle 2025;16(2):e13804
Heavy enough is the whole question, and heavy enough changes from one day to the next.
The AI-based strength measurement system reads exactly what level of effort you are putting in on the day, so the load in your hands is heavy enough to build, every session.
04 · Strength and conditioning
Aerobic capacity and muscle both predict how long you live. The research shows lifting builds both, and cardio on its own costs you muscle when you are losing weight.
A review of 37 studies in 1,641 healthy older adults found resistance training alone raised peak oxygen uptake, which is the ceiling on how much oxygen your body can use when it works hard. The effect was clearest in programs shorter than 24 weeks.
160 older adults with obesity followed the same diet for six months. The aerobic group lost about 5% of its lean mass. The lifting group lost about 2%. Everyone lost a similar amount of weight. The group that did both improved its physical performance the most.
Villareal et al., New England Journal of Medicine 2017;376(20):1943–1955
Conditioning here is built into the lifting. Resting heart rate is one of the six markers we retest, so you see whether it worked.
Every study on this page is published and peer reviewed, and every one is linked to the paper itself. None of them were funded by us or written by us. The plain-English version of what it all means is on The Science, and the long version of each topic is in the Lab Notes.
You have never had any of this measured. That is the whole point.
Request Your Longevity Analysis