What is the actual evidence for longevity science? Cardiorespiratory fitness, grip strength, and the gap between association and cause

TL;DR The hardest longevity data there is are all "associations", not causes — and that one sentence is what this article most wants you to take away. The association between cardiorespiratory fitness (CRF) and mortality is startlingly large: Mandsager et al. (JAMA Netw Open, 2018) followed 122,007 patients who had undergone exercise treadmill testing over a median 8.4 years, with 13,637 deaths, and the risk-adjusted hazard ratio (HR) for the elite performers against the lowest group was 0.20 (95% CI 0.16–0.24); Kokkinos et al. (JACC, 2022), using 750,302 US veterans over a median 10.2 years with 174,807 deaths, found an HR of 4.09 (95% CI 3.90–4.20) for the least fit 20% against the extremely fit, and the paper itself writes that being unfit carried a greater risk than any of the cardiac risk factors it examined [Note 1]. On grip strength, Leong et al. (Lancet, 2015, the PURE study) found in 139,691 people across 17 countries, over a median 4.0 years with 3,379 deaths, an all-cause mortality HR of 1.16 (95% CI 1.13–1.20) per 5 kg lower grip strength. But — push those numbers up, and does mortality follow them down? No trial has shown it. Generation 100 (BMJ, 2020) was a 5-year randomised controlled trial in 1,567 residents of Trondheim, Norway, aged 70–77 (the paper describes its sampling frame as the general population of older adults in Trondheim, Norway — one city's older population, not a national sample). Control-group mortality was 4.7%; the high-intensity interval group showed an absolute risk reduction of 1.7 percentage points (HR 0.63, 95% CI 0.33–1.20) and the moderate-intensity group an absolute increase of 1.2 percentage points (HR 1.24, 95% CI 0.73–2.10) — and every one of those confidence intervals crosses 1. ⚠️ But note equally: the control group was not sedentary; they too were advised to follow the physical activity guidelines then in force, so this trial cannot be read backwards as "exercise cannot reduce mortality". The PURE paper's own concluding sentence says that whether improving strength reduces mortality still needs testing. Zone 2 is not as settled as you think. The narrative review by Storoschuk et al. (Sports Med, 2025) states plainly that current evidence does not support Zone 2 training as the optimal intensity for improving mitochondrial or fatty acid oxidative capacity [Note 2], and argues for prioritising higher intensities where training volume is lower. Bryan Johnson appears here only as a way in, not as an authority. As at 3 August 2026, the words rapamycin, lithium and NDGA appear not once on his own public protocol site (the whole site is a single page, counted word by word, with other strings on the same page used as positive controls) — but his other website answered the crawl with HTTP 429, so this is "his public page does not say so", not "he is not taking them". One safety signal few people mention: for chaparral (Larrea tridentata), the plant source of NDGA, the US National Library of Medicine's LiverTox records more than two dozen published cases of liver injury, some progressing to acute liver failure and needing emergency transplantation, with a likelihood score of B — and it states expressly that chaparral preparations have not been shown to be effective in any medical condition [Note 3]. This article will not tell you how to take any drug, will not recommend any supplement or course of treatment, will not recommend any clinic or laboratory, and will not say "take this and you will live longer". Individual drugs, doses and exercise prescriptions are for a doctor to decide.

#How long do people in Hong Kong already live? Get the baseline right first

Hong Kong's life expectancy really is at the very top, but the Hong Kong Government's own way of saying so has not been consistent — from "the longest-living city in the world" to "one of the longest-living places in the world" to "among the longest in the developed economies", and those three formulations do not cover the same ground.

Census and Statistics Department interactive data table 115-01011 (statistical item: "expectation of life at birth"; the source field names the Demographic Statistics Section (one) of the Census and Statistics Department and the Statistics Unit of the Department of Health; table data release date 30 June 2026) publishes the following:

• 2025: men 83.3 years, women 88.7 years — marked with a "p", which the department's own symbol table defines as a provisional figure.
• 2024: men 82.7 years, women 88.2 years.
• 2023: men 82.5 years, women 88.1 years.
• 2022: men 80.7 years, women 86.8 years — marked with a "‡", which the department's Chinese symbol note explains as the mortality rate in 2022 having been unusually high during the 2019 coronavirus disease epidemic, so the figures for that year should be interpreted with care [Note 4].
• 2021: men 83.2 years, women 87.9 years.

The scale of that improvement is on the record. Paragraph 2.4 of Hong Kong Life Tables 2016–2046 (2022 edition, published August 2023) states that expectation of life at birth for men rose from 67.8 years in 1971 to 83.2 years in 2021, an increase of 15.5 years over 50 years; and for women from 75.3 years in 1971 to 87.9 years in 2021, an increase of 12.6 years over 50 years [Note 5].

In one sentence: within half a century men in Hong Kong gained 15.5 years of life and women 12.6 — and all of it happened before any anti-ageing drug or any biological age test existed.

The gap between the sexes has an official projection too. Paragraph 2.6 of the same document: the gap narrowed from 7.5 years in 1971 to 5.7 years in 2019, fluctuated slightly during the epidemic years 2020 to 2022, and is projected to stay at around 5.4 to 5.5 years from 2023 to 2046 [Note 5].

Do the arithmetic (calculated here): on the provisional 2025 figures above, 88.7 − 83.3 = 5.4 years, exactly at the lower end of the 5.4–5.5 range that document projects. This is a subtraction performed here on figures already quoted, not a gap figure published by the department.

⚠️ One discrepancy to note. Table 24 of Hong Kong Life Tables 2016–2046 prints 81.3 / 87.2 for 2022 (provisional); the current interactive data table 115-01011 gives 80.7 / 86.8 for 2022. The same year, two government figures — revised after publication. This article cites the interactive data table throughout.

The same fact, three government formulations: the difference is itself worth knowing. Paragraph 100 of the Chief Executive's 2022 Policy Address is the broadest — "Hong Kong has the longest life expectancy in the world"; paragraph 146 of the same report also says it is the longest-living place in the world. A year later it narrowed: paragraph 113 of the 2023 Policy Address becomes "Hong Kong has among the longest life expectancies in the world". And the Census and Statistics Department's Statistics and Us (November 2025) uses a third and narrowest formulation: the expectation of life at birth for men and women in Hong Kong is among the longest in the developed economies [Note 6].

In one sentence: "the longest-living city in the world", "one of the longest-living places in the world" and "among the longest in the developed economies" are three statements of different scope — first in the world, near the top in the world, and near the top within the developed economies. The department's sentence is the narrowest of the three and the only one written by the body that produces the figures.

⚠️ For anyone who hears, or wants to quote, "Hong Kong has the longest life expectancy in the world": that sentence does have Hong Kong Government backing — paragraph 100 of the 2022 Policy Address says exactly that. But two things are worth knowing before quoting it: (i) the department that produces the figures uses a narrower formulation itself; and (ii) the Policy Address itself changed in 2023 from "city" plus "longest" to "place" plus "among the". One government, three formulations — and which sentence you pick decides how large a claim you are making.

#How did this "longevity science" market grow up?

The driver is not a new drug. It is demographics — and the lengthening tail between healthspan and lifespan.

Table 1 of Hong Kong Population Projections 2022–2046 (August 2023, published in Chinese and English together) projects the share of the population aged 65 and above rising from 20% in 2021 to 34% in 2046; the elderly dependency ratio from 282 to 560; and the median age from 46.3 to 53.4. Paragraph 2.4 of the same document uses a different basis: excluding foreign domestic helpers, the elderly population rises from 1.45 million in 2021 (20.5% of the total population) by 1.29 million to 2.74 million in 2046 (36.0%); compared with the increase of about 0.8 million over the preceding 25 years (1996 to 2021), the pace of increase in the elderly population is clearly accelerating [Note 7].

⚠️ 20% and 20.5% do not share a denominator. Table 1's 20% (and 34% for 2046) includes foreign domestic helpers; paragraph 2.4's 20.5% (and 36.0%) excludes them. The two figures cannot be compared inside one sentence.

The actual figures from Census and Statistics Department interactive data table 110-01004 fill in the recent years: the elderly dependency ratio (defined as the number of people aged 65 and above per thousand people aged 15 to 64) rose from 282 in 2021 to 356 in 2025; and the median age of the population in 2025 was 48.4 (men 48.0, women 48.6).

A longer life is not the same as a longer healthy life — and the Government has said so. Page 3 of the Centre for Health Protection's Non-Communicable Diseases Watch of February 2015, in the piece on ageing well, states that people at higher health risk develop disability earlier in later life and have more of it; that conversely, adopting a healthier lifestyle not only extends life but lengthens the time spent in health; and that compared with the high-risk group, the low-risk group's onset of disability can be postponed by more than five years [Note 8].

Which is to say: the Government's own position is that behavioural factors affect not only when you die but when you start to fail. That distinction — lifespan is not healthspan — is exactly where the whole longevity industry makes its pitch.

Page 3 of the same publication also cites Hong Kong's own comorbidity figures: a thematic household survey conducted between October 2011 and January 2012 found that nearly three quarters (73.7%) of people aged 65 and above had a chronic disease diagnosed by a Western medicine practitioner, including hypertension (46.0%), diabetes (20.0%), heart disease (9.8%), cancer (3.5%) and stroke (3.2%) [Note 8].

⚠️ That publication dates from 2015, and the population projection it carries (30.2% aged 65 and above in 2041) comes from a round of projections since superseded. This article uses only its qualitative statements; every population figure here comes from the 2023 round.

⚠️ If you think "longevity tech" appeared only in the last three years: the demand-side curve — 65 and above going from 20% towards 34%, the elderly dependency ratio climbing from 282 to 356 in four years — was drawn a long time ago. The supply side following that curve is commercial logic, not a scientific breakthrough.

#How strong is the association between cardiorespiratory fitness and mortality?

Stronger than smoking, diabetes and coronary artery disease — but of those three studies, two (Mandsager, Kokkinos) are observational cohorts and the third (Kodama) is in fact a meta-analysis pooling 33 studies rather than a cohort in its own right; and all of their subjects were people referred for clinical exercise testing.

First, the terms. Cardiorespiratory fitness (CRF) is the body's capacity to deliver and use oxygen during sustained exercise, estimated clinically on a treadmill test and usually expressed in METs (metabolic equivalents) — 1 MET is roughly the oxygen consumption of sitting still. VO2max (maximal oxygen uptake) is another way of expressing the same thing. All of the studies below are in English-language journals with no Chinese original, and their original sentences are set out in the notes at the end.

First: the largest single treadmill cohort to date. Mandsager et al. (JAMA Network Open, 2018;1(6):e183605) — patients at a tertiary care academic medical centre in the United States between 1 January 1991 and 31 December 2014, 122,007 people (mean age 53.4, 59.2% male), median follow-up 8.4 years, 13,637 deaths over 1.1 million person-years. The result: risk-adjusted all-cause mortality was inversely proportional to cardiorespiratory fitness and lowest in elite performers; elite against low gave an adjusted hazard ratio of 0.20 (95% CI 0.16–0.24), and elite against high 0.77 (95% CI 0.63–0.95) [Note 9].

The same paper set this against traditional risk factors: low fitness against elite gave an adjusted hazard ratio of 5.04 (95% CI 4.10–6.20); below average against above average 1.41 (95% CI 1.34–1.49). And in the same models, coronary artery disease was 1.29, smoking 1.41 and diabetes 1.40 [Note 9].

In one sentence: within this cohort, the mortality risk carried by "fitness below average" is the same order of magnitude as the risk carried by smoking.

Second: the one with the largest denominator. Kokkinos et al. (J Am Coll Cardiol, 2022;80(6):598-609) — 750,302 US veterans aged 30 to 95 (mean age 61.3), median follow-up 10.2 years over 7,803,861 person-years, 174,807 deaths, an average of 22.4 per 1,000 person-years. The lowest mortality risk was at about 14.0 METs (men HR 0.24, 95% CI 0.23–0.25; women HR 0.23, 95% CI 0.17–0.29); the least fit 20% against the extremely fit gave HR 4.09 (95% CI 3.90–4.20). The paper's conclusion: being unfit carried a greater risk than any of the cardiac risk factors examined [Note 1].

⚠️ This cohort is US veterans, and specifically people referred for a clinical exercise treadmill test — not a general population, and not Hong Kong people. The paper itself offers no statement transferable to Hong Kong.

Third: the one that converts into "what is 1 MET worth". Kodama et al. (JAMA, 2009;301(19):2024-35) — 33 studies, with 102,980 participants and 6,910 cases in the all-cause mortality analysis. Per 1 MET higher, the pooled relative risk for all-cause mortality was 0.87 (95% CI 0.84–0.90); for cardiac and cardiovascular events 0.85 (95% CI 0.82–0.88). Low fitness against high fitness gave an all-cause mortality RR of 1.70 (95% CI 1.51–1.92). The paper defines its three groups as low fitness below 7.9 METs, intermediate 7.9–10.8 METs, and high 10.9 METs or above [Note 10].

Do the arithmetic (calculated here, using two figures from that same paper): from the top of the "low" band at 7.9 METs to the bottom of the "high" band at 10.9 METs is a difference of 3 METs. At a relative risk of 0.87 per MET, a gap of 3 METs corresponds to 0.87³ = 0.659, and inverted, 1 ÷ 0.659 = about 1.52 times. The paper's own direct low-against-high comparison reports 1.70 times. The two point the same way and are the same order of magnitude, the difference coming from how wide the bands themselves are. This is an exponentiation performed here on figures already quoted, not a figure published by Kodama et al. — and either way it remains an association, not "raise your METs by 3 today and cut your mortality risk by 30% to 40%".

⚠️ If you have had a health check, your report carries a VO2max or a MET figure, and you do not know how to read it: these numbers are prognostic markers — how well they predict, and what happens if you change them, are two completely different questions. The next section takes the second.

#Why is grip strength related to mortality, and what is it actually measuring?

Grip strength is a simple and inexpensive method of risk stratification (the paper's own wording), and in the PURE study it predicted death better than systolic blood pressure — but the researchers state in writing that whether raising grip strength lowers mortality has never been tested.

Leong et al. (Lancet, 2015;386(9990):266-73) — the PURE (Prospective Urban Rural Epidemiology) study, 17 countries of differing income and sociocultural background, recruiting 142,861 people between January 2003 and December 2009, of whom 139,691 with known vital status entered the analysis; median follow-up 4.0 years (IQR 2.9–5.1), 3,379 deaths (2%), measured on a Jamar dynamometer. The results: per 5 kg lower grip strength, the hazard ratio was 1.16 (95% CI 1.13–1.20) for all-cause mortality, 1.17 for cardiovascular death, 1.17 for non-cardiovascular death, 1.07 for myocardial infarction and 1.09 for stroke; and grip strength predicted all-cause and cardiovascular mortality better than systolic blood pressure did [Note 11].

What it did not predict matters just as much. The same passage states that no significant association was found between grip strength and incident diabetes, the risk of hospital admission for pneumonia or COPD, injury from a fall, or fracture [Note 11].

Which is to say: grip strength is not an all-purpose health score. In this study it had no significant association with fall injury or fracture — the two things many people assume grip strength is measured to predict.

And the dose relationship for muscle-strengthening activity itself is not "more is better". Momma et al. (Br J Sports Med, 2022;56(13):755-763) pooled 16 cohort studies: muscle-strengthening activity was associated with a 10–17% lower risk of all-cause mortality, cardiovascular disease, total cancer, diabetes and lung cancer; but showed no association with colon, kidney, bladder or pancreatic cancer. All-cause mortality, cardiovascular disease and total cancer showed J-shaped associations, with the maximum risk reduction (approximately 10–20%) at approximately 30 to 60 minutes a week of muscle-strengthening activity [Note 12].

The paper's concluding sentence adds one thing itself: given the J-shaped associations observed, the influence of a higher volume of muscle-strengthening activity on all-cause mortality, cardiovascular disease and total cancer is unclear [Note 12].

A contrast: the World Health Organization recommends muscle-strengthening activity on "two days or more each week" for adults (quoted by the Centre for Health Protection below); while the maximum risk reduction Momma et al. observed falls at approximately 30 to 60 minutes a week. The two do not conflict — one is a frequency and the other a total duration — but together they say this: the dose threshold is low, and the return on adding more is unclear.

⚠️ If you are over 60, or have an older person at home, and want to know what measuring grip strength is for: it is a cheap, quick, repeatable risk-stratification tool (the PURE paper's own words are "simple, inexpensive risk-stratifying method"). But stratification is not diagnosis, and it is not the same as changing that number changing the outcome.

#The association is this strong — so why still not say "raise your VO2max and you will live longer"?

Because the studies that measure these numbers and the studies that change them are two entirely different bodies of work — and the second either has not been done, or was done and came out null. This is the most important section here.

First, separate three things that are easily run together:

• Prognostic association: fitter people have lower mortality. Everything in the previous section is this.
• Observational association between a change in exposure and an outcome: people who go from unfit to fit have lower mortality than people who stay unfit. This is still observational, because nobody was randomised into becoming fit.
• The causal effect of an intervention: randomise a person into exercising, and does their mortality fall. Only a randomised controlled trial can answer this.

The second layer has data, and handsome data. Blair et al. (JAMA, 1995;273(14):1093-8) — 9,777 men who had two preventive medicine examinations, with 223 all-cause deaths and 87 cardiovascular deaths. Men unfit at both examinations had the highest age-adjusted all-cause death rate, at 122.0 per myriad man-years; men fit at both had the lowest, at 39.6; men who went from unfit to fit had 67.7, a 44% reduction in mortality risk (95% confidence interval 25% to 59%) relative to those who remained unfit [Note 13].

⚠️ But this is still not causal evidence: these men became fit of their own accord, not by randomisation; and they were all men, all self-referred attenders at a preventive medicine clinic. "Healthy to begin with, therefore able to exercise" and "exercised, therefore healthy" cannot be separated in this design.

The third layer has been done, and it was null. Stensvold et al., the Generation 100 study (BMJ, 2020;371:m3485; ClinicalTrials.gov NCT01666340) — the paper describes its sampling frame as the general population of older adults in Trondheim, Norway, and the authors are mainly from the Norwegian University of Science and Technology (NTNU) and St Olav's University Hospital; of 6,966 people born between 1936 and 1942, 1,567 (790 women, mean age 72.8) were randomised, for five years:

• High-intensity interval training (HIIT, about 90% of maximum heart rate), two sessions a week, n = 400
• Moderate-intensity continuous training (MICT, about 70% of maximum heart rate), two sessions a week, n = 387
• Control: follow the national physical activity guidelines, n = 780

The result: there was no difference in all-cause mortality between the control group and the combined MICT and HIIT group. Taking the control group as reference (observed mortality 4.7%), an absolute risk reduction of 1.7 percentage points was observed after HIIT (HR 0.63, 95% CI 0.33–1.20) and an absolute increased risk of 1.2 percentage points after MICT (HR 1.24, 95% CI 0.73–2.10). With MICT as the reference group, HIIT showed an absolute risk reduction of 2.9 percentage points for all-cause mortality (HR 0.51, 95% CI 0.25–1.02). The conclusion reads: combined MICT and HIIT has no effect on all-cause mortality compared with recommended physical activity levels; however, a lower all-cause mortality trend was observed after HIIT compared with controls and MICT [Note 14].

⚠️ Two things have to be said together; leaving either out misleads.

(i) Every confidence interval crosses 1. 0.63 (0.33–1.20), 1.24 (0.73–2.10), 0.51 (0.25–1.02) — all three contain 1, meaning the possibility of "no difference" has not been excluded.

Do the arithmetic (calculated here, on the figures already quoted): control group 780 × 4.7% ≈ 37 deaths; HIIT group 400 × (4.7% − 1.7%) = 3.0% ≈ 12; MICT group 387 × (4.7% + 1.2%) = 5.9% ≈ 23. Roughly 72 deaths across the whole trial in five years. That is the direct reason the confidence intervals are this wide — 1,567 people, five years, some seventy events cannot statistically separate a difference of moderate size. These are multiplications performed here on percentages and group sizes already quoted, not death counts published by the study.

(ii) The control group was not "no exercise". The paper states it itself: control participants chose to perform more of their physical activity as HIIT than the participants in the MICT group did, which meant the controls achieved an exercise dose at an intensity between the MICT and HIIT groups [Note 14]. Add that at baseline 87.5% of participants self-reported good general health and 80% self-reported a moderate or high level of physical activity.

So this trial compares "supervised structured exercise" with "following the national guidelines on your own" — not "exercise" with "no exercise". Anyone who writes up Generation 100 as "exercise cannot extend life" has misread that control group.

⚠️ One more limit of scope to remember: this trial's sampling frame is the older population of one city (Trondheim), not a national Norwegian sample, and certainly not Hong Kong people. At baseline 87.5% of participants self-reported good general health — that is, this group started out healthier than their contemporaries at large.

On grip strength the authors put it more directly still. The last sentence of the PURE paper's conclusion states in writing that further research is needed to identify determinants of muscular strength and to test whether improvement in strength reduces mortality and cardiovascular disease [Note 11].

⚠️ If anyone uses a figure of the form "people with high VO2max have X times lower mortality" to sell you a test, a machine or a package of treatments: that figure is very likely true, and genuinely comes from the papers above; but getting from that figure to "so buying this will make you live longer" crosses a randomised controlled trial that nobody has yet succeeded in doing. That gap is the thing this article most wants you to remember.

#Is Zone 2 the best exercise intensity?

No — at least one peer-reviewed review addressing that claim directly concludes that current evidence does not support it. Popular accounts of this point are generally a great deal more polite.

First, the term. Zone 2 in fitness circles usually means low-intensity continuous exercise below the lactate threshold, popularly described as "the intensity at which you can still speak in sentences". The fashionable claim in recent years is that it is particularly good for mitochondrial function and therefore the "optimal" intensity for longevity.

Storoschuk et al. (Sports Medicine, 2025;55(7):1611-1624; PMID 40560504) is a narrative review addressing that claim head on. It notes that these recommendations stem largely from observational data on elite endurance athletes, who do large volumes of Zone 2 training and possess high mitochondrial and fatty acid oxidative capacity; but the authors challenge the broad endorsement of Zone 2 training for the general public, as it contradicts substantial evidence supporting high-intensity exercise for improving mitochondrial capacity and cardiometabolic health [Note 2]. Their conclusion: current evidence does not support Zone 2 training as the optimal intensity for improving mitochondrial or fatty acid oxidative capacity; further, evidence suggests that prioritising higher intensities (above Zone 2) is critical to maximise cardiometabolic health benefits, particularly in the context of lower training volumes [Note 2].

Three limits have to be given with it, or this becomes using one review to beat down another claim:

1. It is a narrative review, not a systematic review and not a meta-analysis. Its own words are "we challenge" — this is a position argued, not a pooled estimate. This article will not promote it beyond that.
2. The declaration of interests is in the same record: Martin J. Gibala is an advisor to and holds equity in Longevity League, Ltd., a US-based company whose services in part relate to exercise [Note 15]. That declaration has to appear wherever this review does.
3. No randomised controlled trial compares Zone 2 with other intensities on mortality or lifespan. What was searched for this is set out under "What this article does not state" at the end. So this whole argument is being fought on surrogate endpoints — mitochondrial capacity, cardiorespiratory fitness — and not on mortality.

⚠️ If you are following a podcast or an app and putting all your training into Zone 2: the smaller your time budget, the more this review's argument applies to you — it says explicitly "particularly in the context of lower training volumes". But equally: this is an argument about mitochondria and cardiorespiratory fitness, not about which way of running makes you live longer, because there is no trial data on the second.

#How much exercise do the official pages tell you to do — and who is actually saying it?

The World Health Organization is saying it, not the Department of Health. The Centre for Health Protection is relaying it.

The introduction to the Centre for Health Protection's "Physical Activity" health information page (page dated 15 April 2025) states that what follows is the World Health Organization's recommendation for adults aged 18 and above (including those living with a chronic condition or disability): they should undertake regular physical activity; they should do at least 150–300 minutes of moderate-intensity aerobic physical activity a week, or at least 75–150 minutes of vigorous-intensity aerobic physical activity, or at least an equivalent combination; they should also do muscle-strengthening activities at moderate or greater intensity targeting all major muscle groups on two days or more each week; and older people aged 65 and above should also, on 3 days or more each week, do varied multicomponent physical activity at moderate or greater intensity emphasising balance and strength training, to enhance functional capacity and prevent falls [Note 16].

In one sentence: aerobic and strength are two separate requirements, not a choice between them; and for people over 65 there is a third (balance). Those three correspond exactly to the three measurements above — cardiorespiratory fitness, grip strength and muscle, and fall risk.

(The academic source of the guideline itself: Bull FC et al., Br J Sports Med, 2020;54(24):1451-1462.)

Local attainment, from the same page: the Behavioural Risk Factor Survey of 2023 found that 14.8% of people aged 18 and above were insufficiently physically active, 16.0% of women and 13.4% of men [Note 16].

⚠️ A fact worth holding on to: that 14.8% is the proportion who are insufficiently active, not the proportion meeting the target, and not a figure from any earlier population health survey. Percentages from two different surveys are not interchangeable.

⚠️ A finding at the level of language, noted in passing. The English version of that same Centre for Health Protection page currently carries two typographical errors — "to prevent fails" (for falls) and "moderate-intensity aerobic physical to more than 300 minutes" (missing activity). The Chinese version has neither problem. This article follows the project's convention of quoting the Chinese original.

Do the arithmetic (calculated here): 150 minutes ÷ 5 days a week = 30 minutes a day; 300 minutes ÷ 5 days = 60 minutes a day. Over 3 days a week instead, that is 50 minutes to 100 minutes a day. This is a division performed here on the 150–300 minutes quoted above, not a daily target published by the Centre for Health Protection or the WHO.

⚠️ If you often hear "the Department of Health says 150 minutes": the source is the WHO and the Centre for Health Protection is quoting it — this is not pedantry: knowing the source is the WHO tells you the figure comes from an international guideline of 2020, whose scope and revision cycle are quite different from those of a local department's own advice.

#What does Bryan Johnson's protocol actually say? (And what it does not)

His role here is as a way in — a public case that makes all the arguments above concrete. He is not an authority on any medical claim, and his own website says as much.

Bryan Johnson is a former American technology entrepreneur who has made his own body a public experiment and published his whole protocol online. This article states only what he has published, and does not recommend that any reader adopt any part of it.

First: his own statement of limits is far more useful than the protocol itself. The foot of protocol.bryanjohnson.com (retrieved 2 August 2026) states that the protocol encompasses a mix of on-label, off-label and unlicensed therapies, as well as research-use-only tests; that some of these tests and therapies are still under scientific investigation and have not received on-label licensing for specific health conditions; that all tests and therapies, whatever their licensing status, carry risks; and that this protocol represents an experimental clinical research project [Note 17].

In one sentence: the publisher states himself that this is an experimental project assessed for one person, containing unlicensed therapies. That is the single most quotable sentence in this whole article.

Second: on the claim that "low-dose lithium and NDGA were added in 2026" — this article found nothing to support it.

As at 2 August 2026, protocol.bryanjohnson.com is a single page for the whole site (its sitemap lists one URL), and this article counted the full text of that page word by word:

The "Rx / Prescriptions" section of that same page lists five prescription medicines: acarbose, tadalafil, candesartan, Jardiance and Repatha; plus three marked as medication for existing conditions (two thyroid medicines, with hypothyroidism stated as diagnosed at 21; and one for hair growth).

This article does not reproduce the doses listed on that page. The reason is not that they are disputed, but that a personal medication list assessed for one individual, reproduced together with its doses, turns in a reader's hands from a piece of reporting into a prescription to copy. Those doses are visible to anyone on that public page; this article's choosing not to restate them is itself part of what it is saying to the reader.

⚠️ Equally important is what that zero does and does not mean. It means "these three are not listed on his published protocol page" — not "he is not taking them". His other domain, blueprint.bryanjohnson.com, answered this crawl (root, product list, and a direct curl request) with HTTP 429 (local_rate_limited) continuously for about 15 minutes on 2 August 2026, and was not examined at all. So this is a bounded "not found", not an established "not there".

Third: his first-hand account of stopping rapamycin. blueprint.bryanjohnson.com/blogs/news/i-stopped-taking-rapamycin (publication date shown on the page 01.25.2025, author Bryan Johnson, obtained by WebFetch on 2 August 2026) states that despite the immense potential from pre-clinical trials, he and his team concluded that the benefits of lifelong dosing of rapamycin do not justify the hefty side effects — intermittent skin and soft tissue infections, lipid abnormalities, glucose elevations, and increased resting heart rate [Note 18].

⚠️ Be careful with the date. The post says "On September 28th" and gives no year; the post itself is dated 25 January 2025. The same post then says "Additionally, on October 25th, a new pre-print…", and that preprint (see the next paragraph) was posted in October 2024 — so on the post's own sequence, 28 September precedes the preprint, making it 2024. But that is an inference drawn here from the post's internal order, not a year the post writes out, and this article therefore does not treat "28 September 2024" as a stated fact.

Fourth: on "16 biological age clocks show rapamycin accelerates ageing" — the preprint was found, but it establishes less than the popular version claims.

The body of that post says only "a new pre-print", without naming a title or an author in the sentence; but the post carries a "Sources" list at its foot, the sentence is followed by the numeral [5], and item 5 of that list is the preprint's URL: pubmed.ncbi.nlm.nih.gov/39484592/. Checked item by item, that preprint is:

• Title: "DNAm aging biomarkers are responsive: Insights from 51 longevity interventional studies in humans"
• Authors: Sehgal R, Borrus D, Kasamato J et al. (13 in all)
• Platform: bioRxiv (a preprint server), version marked posted October 27, 2024 (the date on the PubMed record is 25 October 2024)
• DOI: 10.1101/2024.10.22.619522; PMID 39484592

Its abstract states that the authors curated TranslAGE-Response, a harmonised database of 51 public and private longitudinal interventional studies, and calculated a consistent set of 16 prominent epigenetic clocks for each study, along with 95 other DNA methylation biomarkers that help explain changes in each clock [Note 19].

So the "16 clocks" figure has a source, and rapamycin is indeed one of the interventions the preprint covers (its figures list "Rapamycin dose 1" through "dose 3" and file rapamycin under "Pharmacological").

⚠️ But these three points must all be given together:

1. It is a preprint and has not been peer reviewed. Every page of it carries the statement "which was not certified by peer review".
2. The directional conclusion that "rapamycin accelerates ageing" is Johnson's description of the preprint, not a sentence the preprint's own text writes. Its text discusses the strength of response of different classes of intervention on DNA methylation biomarkers, and has no sentence singling out rapamycin as accelerating ageing; direction has to be read off the effect values in the figures. This article did not obtain those item-by-item effect values, and therefore restates no directional conclusion.
3. The other source popularly cited for this claim (PMC12226543) is the wrong one. Counted word by word, that article contains "epigenetic" 0 times and "clock" 0 times — it is in fact a narrative review by Roark and Iffland in Front Aging in 2025, with nothing to do with epigenetic clocks.

In one sentence: the preprint is real, the "16 clocks" is real, and rapamycin being among the interventions analysed is real; but the words "accelerates ageing" rest at present on Johnson's restatement alone, with no original sentence this article can check against.

Fifth: the sauna figure in circulation is wrong at source, and this article does not inherit it. The "sauna" section of Johnson's page says that sauna use "reduce cardiovascular mortality by 63%". The corresponding original study is Laukkanen et al. (JAMA Intern Med, 2015;175(4):542-8) — 2,315 men aged 42 to 60 in eastern Finland, median follow-up 20.7 years. In that study:

• the hazard ratio of 0.37 (95% CI 0.18–0.75) is for sudden cardiac death, not for cardiovascular mortality;
• the reference group is men who used a sauna once a week, not men who never used one (1 a week 601 men, 2–3 times 1,513, 4–7 times 201);
• and on session length: more than 19 minutes against less than 11 minutes gave a hazard ratio for sudden cardiac death of 0.48 (95% CI 0.31–0.75), with significant inverse associations for fatal coronary heart disease and fatal cardiovascular disease too, but no significant association with all-cause mortality.

In one sentence: this is an observational association in a cohort of middle-aged Finnish men, and it compares "more sauna" with "less sauna", not "some" with "none". The popular write-up usually gets two things wrong, and this article corrects both.

Sixth: two widely circulated figures for which this article found no first-hand source, and does not use. (i) "USD 2,000,000 spent a year" — protocol.bryanjohnson.com (recounted word by word on 3 August 2026) contains only two $ characters in its entire text, both in the same sentence about a water filter: an "alkaline water filtration system" marked $1,300, with the same sentence noting that similar systems can be had for $300. There is no other sum anywhere on the page. bryanjohnson.com says only "Two years and millions were spent developing an algorithm". Neither supports the figure of USD 2,000,000 a year. (ii) "About 80% of the effect can be reproduced for under USD 100 a month", attributed to Johnson himself — this crawl found no such sentence anywhere on his own websites. This is a quantified efficacy claim attached to a named person, and without a source it cannot appear. (iii) A claim that a named doctor is still taking rapamycin — no first-hand source obtained, likewise removed.

On price, all that could be obtained is retail product list prices. On an archived snapshot of his official store (snapshot time 15 May 2026), six retail products were listed at: Longevity Mix $49.00, Essential Capsules $49.00, Advanced Antioxidants $49.00, Extra Virgin Olive Oil $39.00, Omega-3 $39.00 and Collagen Peptides $45.00, with the page also carrying "Free US Shipping on all orders $50+". ⚠️ These are the prices of six retail products as at 15 May 2026, not the cost of the whole protocol, and not prices for August 2026.

On biological age, this article says one thing only. Johnson's own page states that these tests are experimental and intended solely for research purposes, and that they should not replace or supplement any clinical tests recommended by licensed medical professionals [Note 20]. What a biological age test measures, whether two runs give the same number, which suppliers exist in Hong Kong and at what price, is the subject of a separate article and is not opened up here.

⚠️ If you are inclined to treat a public case as "the 2026 consensus": the most reliable and most quotable sentence in this one is his own, "This protocol represents an experimental clinical research project" — one person, one team, one experimental regimen assessed for him individually. n = 1.

#Rapamycin: how far is it from the mice to us?

The mouse data are real and handsome; on the human side, the one randomised controlled trial of any size had a null primary endpoint — and that trial's registered sponsor is itself a company that sells rapamycin.

The legal position in Hong Kong first. Searching the Hong Kong pharmaceutical products registration database (each page of which carries "Last Updated: 31-Jul-2026") for rapamycin returns "RAPAMYCIN IS THE SYNONYM / STANDARD NAME OF : sirolimus" and "Record: 1 to 3 of 3" — three registered products, all RAPAMUNE held by Pfizer (1mg tablet, 0.5mg tablet, oral solution 1mg/ml), all legally classified as Part 1, Schedule 1 & Schedule 3 Poison, all with a sale requirement of Prescription only Medicines, which that database itself defines as medicines which must only be purchased with a prescription in a pharmacy [Note 21].

⚠️ A limit of scope that has to be drawn clearly. This registration database records classification and registration certificate holder, not approved indications. So this article cannot say on the strength of this source that sirolimus is not approved in Hong Kong for anti-ageing use — only that this source publishes no approved indications. (A popular formulation runs "its clinically approved indications do not include extending life or anti-ageing"; that sentence goes beyond what this article's sources support, and has been removed.)

The mouse data, two different measurements, which cannot be mixed.

• Harrison et al. (Nature, 2009;460(7253):392-5): mice were first fed rapamycin at 600 days of age and still had extended median and maximum lifespan; measured as the age at which 90% mortality is reached, the increase was 14% in females and 9% in males.
• Miller et al. (Aging Cell, 2014;13(3):468-77): tested at three times the dose, genetically heterogeneous mice showed median lifespan increases of 23% in males and 26% in females, with maximum lifespan also increased in both sexes.

⚠️ 23–26% and 9–14% are not the same thing: the first is median lifespan at three times the dose; the second is the age at 90% mortality with feeding started late in life. Popular accounts generally quote 23–26% as a single fact without any of those three qualifiers; this article restores all of them.

On the human side: the PEARL trial. Moel et al. (Aging (Albany NY), 2025;17(4):908-936; NCT04488601) — 48 weeks, decentralised, double-blind, placebo-controlled, in healthy adults ageing normally, taking placebo, 5 mg or 10 mg of compounded rapamycin weekly. The primary endpoint was visceral adiposity (DXA scan). The results: adverse and serious adverse events were similar across all groups; visceral adiposity did not change significantly; changes in blood biomarkers remained within normal ranges; lean tissue mass and self-reported pain improved significantly for women using 10 mg; and self-reported emotional well-being and general health also improved for those using 5 mg (each effect size and p value at [Note 22]); and no other significant effects were observed [Note 22].

⚠️ That last sentence — "No other significant effects were observed" — is the one most often dropped when this trial is restated. Without it the preceding items read as a run of positive findings; with it, you can see that those few are all that emerged from a mass of secondary endpoints.

Four things must be said together, or this trial gets read as saying what it never said:

1. The primary endpoint was null. The "improvements" are all secondary or exploratory endpoints — but they cannot all be described as "subgroup findings by sex", nor all as "self-report scales". The two improvements in the 10 mg group (lean tissue mass, self-reported pain) were confined to women; for the two in the 5 mg group (self-reported emotional well-being, self-reported general health) the original notes no sex. And of the four improvements, only three (self-reported pain, self-reported emotional well-being, self-reported general health) really are self-report scales — lean tissue mass is an objective figure measured by DXA scan, not self-reported.
2. The trial's sponsoring body is a company that sells rapamycin. This has three separate records, set out one by one here, because "conflict of interest" and "who ran it" are two different things:
   • The registration record: the "Lead Sponsor" field of ClinicalTrials.gov NCT04488601 is AgelessRx (agency class INDUSTRY), the "Responsible Party" is SPONSOR, and the trial location is also given as AgelessRx (Chicago, Illinois, USA). ⚠️ The same registration record also lists the University of California, Los Angeles (UCLA) as a Collaborator — a point popular restatements generally omit, and which is set out here alongside.
   • The paper's declaration of interests names seven authors as employees and shareholders of AgelessRx [Note 23].
   • The company does sell the drug: AgelessRx's own product page (agelessrx.com/rapamycin/, retrieved 3 August 2026) offers it publicly under the heading "Rapamycin Rx", with the page marked "Starting at $65 / mo" and "Compounded and generic options".
3. The formulation used reaches only about a third of the blood concentration of the commercial product. The full paper states that it was subsequently discovered that compounded rapamycin did indeed have approximately a third of the concentration in blood after 24 hrs relative to commercial, and that while doses are listed at the advertised compounded dose, equivalent effective doses for compounded forms are approximately 66% less [Note 24].
4. Size and duration: 114 people completed the study (5 mg group 40, 10 mg group 35, placebo 39), with a further 11 withdrawing and not included in the analysis; 48 weeks. Do the arithmetic (calculated here): 114 + 11 = 125 randomised; serious adverse events 1 + 2 + 3 = 6 (1 in the 10 mg group, 2 in the 5 mg group, 3 on placebo); non-serious adverse events 117 + 116 + 122 = 355. These are additions performed here on the itemised figures the paper lists, not totals the paper writes out itself.

In one sentence: PEARL is a safety trial, not evidence of life extension. The paper's own concluding sentence says that future work will evaluate the benefits of a broader range of rapamycin doses on healthspan metrics for longevity, and will aim to more comprehensively establish efficacy [Note 22].

Incidentally, PEARL also ran an epigenetic age sub-study (a sample of 24), and the result was that within the epigenetic testing results they saw no meaningful significant changes between groups [Note 22].

How the professional literature describes healthy people taking rapamycin for ageing. Roark and Iffland (Front Aging, 2025;6:1628187, CC BY) state that the long-term effects and safety of chronic mTOR inhibition in healthy humans, and whether rapamycin can truly "slow" human ageing or prevent age-related diseases without unacceptable side effects, are unknown; and that as rapamycin gains popularity for its anti-ageing potential, online longevity clinics have emerged offering access to the drug with minimal medical oversight, a semi-regulated availability that raises ethical concerns regarding patient safety, misinformation and the potential for serious harm [Note 25].

⚠️ If you are considering getting rapamycin online or from a clinic that will not follow you up: in Hong Kong it is a Part 1, Schedule 1 and Schedule 3 poison and a prescription-only medicine — which means the lawful route necessarily includes a registered doctor who takes responsibility for following you up. The 23–26% in mice was median lifespan at three times the dose; on the human side there is one 48-week safety trial in 114 people with a null primary endpoint. The distance between those two things is not "a few years"; it is one trial that has never been done.

#Micro-dose lithium and NDGA: are these two worth touching?

Neither has a human longevity trial. And one of them carries a liver safety signal that is very rarely mentioned.

#Lithium: the drinking water data do not point one way

One URL popularly cited for "areas with lithium in the drinking water have better cognition" (a nature.com link), crawled on 2 August 2026, has the title "Integration of optogenetics with complementary methodologies in systems neuroscience" — a methodological review of optogenetics, with nothing to do with lithium. That citation does not stand, and this article does not inherit it.

The actual drinking-water literature is Kessing et al. (JAMA Psychiatry, 2017;74(10):1005-1010) — a Danish nationwide nested case-control study of 73,731 patients with dementia against 733,653 controls. A nonlinear association was observed: relative to those exposed to 2.0 to 5.0 µg/L, the incidence rate ratio for dementia was 0.83 (95% CI 0.81–0.85) in those exposed to more than 15.0 µg/L, 0.98 (95% CI 0.96–1.01) at 10.1 to 15.0 µg/L, and 1.22 at 5.1 to 10.0 µg/L (95% CI 1.19–1.25) [Note 26].

In one sentence: it is not "the more lithium the better". One of the exposure bands (5.1–10.0 µg/L) corresponds to a dementia incidence rate ratio that is 22% higher. Other researchers published commentary in the same journal after that paper appeared; this article did not obtain them individually and so gives no count. Any account that turns this body of data into "lithium protects the brain" has left out the reversed band inside its own source.

The high-profile Nature paper of 2025 is mice plus human tissue, not a human trial. Aron et al. (Nature, 2025;645(8081):712-721) found lithium significantly reduced in the brains of patients with mild cognitive impairment, and, on reducing endogenous cortical lithium by about 50% in mice, saw increased amyloid deposition and tau phosphorylation; supplementation with lithium orotate in mice prevented the pathological changes and the memory loss. The paper's own wording is that lithium replacement with amyloid-evading salts is a potential approach to the prevention and treatment of Alzheimer's disease [Note 27].

⚠️ The easiest way to misread this paper is as a licence to go and buy lithium orotate. It is a mouse model plus post-mortem and tissue analysis, with no human trial.

What is known about lithium toxicity — but note which dose it measures. McKnight et al. (Lancet, 2012;379(9817):721-8), a systematic review and meta-analysis of 385 studies: glomerular filtration rate reduced by 6.22 mL/min on average; 18 of 3,369 people (0.5%) on renal replacement therapy; an odds ratio for clinical hypothyroidism against placebo of 5.78 (95% CI 2.00–16.67); thyroid stimulating hormone raised by 4.00 iU/mL on average; blood calcium raised by 0.09 mmol/L; parathyroid hormone raised by 7.32 pg/mL; an odds ratio for weight gain of 1.89 (95% CI 1.27–2.82). ⚠️ A limit of scope that must be stated: this is the toxicity picture at therapeutic doses in patients with mood disorders, not at micro-doses. It cannot be used to say micro-dose lithium is dangerous, nor to say it is safe — it is about what is already known of the drug itself.

As for the "supplement-grade" lithium on the market, this article found only one human case report. Pauzé and Brooks (J Med Toxicol, 2007;3(2):61-2): an 18-year-old woman deliberately took 18 capsules of an online "dietary supplement" containing lithium, each labelled as 120 mg of lithium orotate; her serum lithium about 90 minutes later was 0.31 mEq/L, and nausea and tremor resolved after 3 hours of observation. The paper also supplies a conversion: each 100 mg of (organic) lithium orotate contains 3.83 mg of elemental lithium; each 100 mg of (inorganic) lithium carbonate contains 18.8 mg.

Do the arithmetic (calculated here, using the two conversion figures above): one 120 mg capsule of lithium orotate ≈ 120 × 3.83 ÷ 100 = 4.6 mg of elemental lithium; 18 capsules ≈ 82.7 mg. For comparison, one 300 mg lithium carbonate capsule of the kind registered in Hong Kong ≈ 300 × 18.8 ÷ 100 = 56.4 mg of elemental lithium. Which is to say that the elemental lithium taken in that case is about 1.5 capsules of a registered prescription lithium product. These are multiplications and divisions performed here on conversion ratios already quoted, not a comparison published by any source. ⚠️ That case is n = 1, a deliberate overdose, with mild and self-resolving symptoms — the weakest tier of evidence. This article cites it because it is the only human safety data on a consumer lithium orotate product that could be found, and because the reader has no other way to look up that conversion ratio.

Legal position in Hong Kong: searching the pharmaceutical products registration database for lithium carbonate returns "Record: 1 to 7 of 7", with all seven products legally classified as Part 1, Schedule 1 & Schedule 3 Poison and all with a sale requirement of Prescription only Medicines. ⚠️ A limit of scope: this establishes that the Hong Kong register of pharmaceutical products carries no over-the-counter or "supplement-grade" lithium carbonate; it does not establish that none can be bought — unregistered products sold as "health food" are by definition outside that register (see the next section), and this article did not search the database under other lithium salt names such as lithium orotate.

#NDGA: a liver signal that is rarely mentioned

NDGA (nordihydroguaiaretic acid) is a compound derived mainly from the shrub Larrea tridentata (the creosote bush, whose preparations are commonly called chaparral).

The animal data stand, but have to be given with their limits. From the US National Institute on Aging's Interventions Testing Program:

• Strong et al. (Aging Cell, 2008;7(5):641-50): pooled data from three sites showed NDGA extended the lifespan of male mice (p = 0.0006). But the same passage also states that when the proportion of mice alive at the age of 90% mortality was used as a surrogate measure of maximum lifespan, neither NDGA (p = 0.12) nor aspirin (p = 0.16) had a significant effect [Note 28].
• Strong et al. (Aging Cell, 2016;15(5):872-84): the original dose, and doses three times higher and three times lower, all reproduced the male-specific median lifespan extension; the paper states that the effects of NDGA were dose dependent and male specific but without an effect on maximal lifespan [Note 28].

In one sentence: male mice, median lifespan, no effect on maximum lifespan. Popular accounts usually write only "extends lifespan in male mice", dropping the last two qualifiers.

⚠️ And the following passage is almost never raised in popular discussion. The "Chaparral" entry in the US National Library of Medicine's LiverTox database states that chaparral extracts have been linked to several cases of clinically apparent liver injury, some of which have led to acute liver failure and the need for emergency liver transplantation [Note 3]. The same entry sets out more: liver injury attributable to chaparral was first reported in 1990, and more than two dozen cases of clinically apparent liver injury have been published since, mostly from the United States, Canada and Australia; the time to onset varied from 3 weeks to several years, but was usually within 3 to 12 weeks of starting daily ingestion or increasing the daily dose; several reported cases have been severe and some have led to emergency liver transplantation, while others appeared to result in cirrhosis; despite the several reports, over-the-counter products containing chaparral remain available commercially and on the internet; and for unclear reasons no case of liver injury clearly implicating chaparral has been published since 2005 [Note 3]. Its likelihood score is B — a likely but now rare cause of clinically apparent liver injury; the same entry also states that recurrence after reexposure has occurred [Note 3].

And on efficacy, LiverTox itself names the anti-ageing selling point and dismisses it in the very next breath. Two consecutive sentences in the same entry's Background section read: chaparral is also claimed to retard aging and aid in wellbeing; however, chaparral preparations have not been shown to be effective in any medical condition [Note 3].

In one sentence: the pitch for this substance and the sentence refuting the pitch appear one after the other inside the same US National Library of Medicine entry.

⚠️ Be precise: the subject of this LiverTox entry is the plant extract chaparral, of which NDGA is only the principal constituent; it is not a safety assessment of purified NDGA at a stated dose, and LiverTox itself notes that liver injury in some cases may have arisen from impurities or improper preparation. So what this article can say is: the plant source of NDGA carries an established hepatotoxicity signal, including records of acute liver failure and emergency transplantation; and that plant preparation has no proven efficacy for any medical condition. This article will not turn that passage into "NDGA will cause liver failure".

Legal position in the United States: NDGA was banned from food there back in 1968. Part 189 of Title 21 of the US Code of Federal Regulations is itself titled Substances Prohibited From Use in Human Food; section 189.165 in its subpart C is headed Nordihydroguaiaretic acid (NDGA), and reads: NDGA is the chemical 4,4′-(2,3-dimethyltetramethylene) dipyrocatechol, C18H22O4, occurring naturally in the resinous exudates of certain plants, with the commercial product synthesised, and formerly used as an antioxidant in foods; and any food containing added NDGA is deemed adulterated in violation of the act, on an order published in the Federal Register of 11 April 1968 (33 FR 5619) [Note 29].

⚠️ This provision and the LiverTox sentence above have to be read together. LiverTox writes that NDGA "has been used as a food additive in low concentrations" — in the past tense; 21 CFR 189.165 is the reason for that past tense. A substance now marketed as a longevity supplement has been barred from food in the United States for nearly sixty years. (Limit of scope: this regulation governs food additives. The United States regulates dietary supplements under a separate regime, and this article has not checked NDGA's position under it.)

Legal position in Hong Kong: the pharmaceutical products registration database was searched under five spellings — nordihydroguaiaretic, nordihydroguaiaretic acid, NDGA, masoprocol (NDGA's drug name) and guaiaretic (a partial string check) — and all five returned "Record: 1 to 0 of 0"; the same session returned "Record: 1 to 100 of 109" for metformin as a positive control, showing the query itself works. Which is to say: as at the database's own "Last Updated: 31-Jul-2026", the Hong Kong register of pharmaceutical products carries no product containing NDGA. ⚠️ The same limit of scope as for lithium above: this establishes that it is not a registered pharmaceutical product, not that it cannot be bought in Hong Kong — unregistered products sold as "health food" are outside that register.

What is in the clinical trial registries? The results from ClinicalTrials.gov in the United States (query date 3 August 2026):

In one sentence: lithium orotate has one registered human trial, not yet recruiting, whose endpoint is Alzheimer's disease rather than longevity; NDGA has nine human trials, all of them cancer treatment, and not one on longevity. So the statement "there are no longevity trials" holds — but it holds in the form "it has been tried in humans, for another disease", not "nobody has ever studied it".

⚠️ One search trap, noted here so as not to hit it again: in the clinical trial registries zone 2 is overwhelmingly an anatomical zone of the flexor tendons of the hand, not an exercise intensity band. The count returned by that string cannot be taken as a count of exercise science trials.

⚠️ If you see anyone online selling micro-dose lithium, lithium orotate, chaparral or NDGA supplements: neither has a human randomised controlled trial with longevity or healthspan as its endpoint (the search scope is in the table above and at the end). Lithium carbonate is a prescription medicine in Hong Kong — that is the only name this article searched in the registration database, and it does not cover other lithium salts such as lithium orotate; NDGA is not on the Hong Kong register of pharmaceutical products, has been barred from food in the United States since 1968, and its plant source has published cases of acute liver failure and transplantation. This article presents neither of these as an "option".

#How far does Hong Kong actually regulate "health products"?

The Government's own position is that safety is already regulated by the existing framework, and that "health claims" as such are not directly regulated at this stage — and "anti-ageing" is not among the six classes of claim that are named.

The written reply of the Secretary for Health, Professor Lo Chung-mau, to Legislative Council question twenty-two on the regulation of health food (5 June 2024) states that "health food" covers many kinds of product of differing composition and character, with no consistent international definition or regulation; that products meeting the definition of "pharmaceutical product" under the Pharmacy and Poisons Ordinance and of "proprietary Chinese medicine" under the Chinese Medicine Ordinance must satisfy requirements of safety, quality and efficacy and be registered before they can be sold; that the Government also regulates the labelling and promotion of products making medical or health claims (including products that are neither pharmaceutical products nor proprietary Chinese medicines) through the Undesirable Medical Advertisements Ordinance; and that at this stage the Health Bureau considers that, from a risk perspective, the existing framework regulates the safety of such products to an appropriate degree, and that strengthening public education and the provision of information is more suitable than directly regulating the health claims of "health food" [Note 30].

The same reply lists the six classes of claim regulated under the Fourth Schedule to the Undesirable Medical Advertisements Ordinance (Chapter 231): (i) preventing, eliminating or treating breast lumps, (ii) regulating the function of the genito-urinary system, (iii) regulating the endocrine system, (iv) regulating the sugar or glucose content of the body, (v) regulating blood pressure, and (vi) regulating blood lipids or cholesterol [Note 30].

In one sentence: not one of those six is "anti-ageing", "life extension", "lifespan" or "biological age".

⚠️ But this cannot be written up as "Hong Kong does not regulate anti-ageing claims at all". What this article can establish is much narrower: anti-ageing claims are not among the six classes listed in the Fourth Schedule. They may still be caught by other parts of the Undesirable Medical Advertisements Ordinance, by the Trade Descriptions Ordinance (Chapter 362), or — if the product is in substance a pharmaceutical product — by the Pharmacy and Poisons Ordinance. This article has not checked those three routes and therefore makes no statement about them.

On the scale of enforcement, the same reply states that over the same period the Department of Health handled 74 cases resulting in conviction for unregistered pharmaceutical products, the heaviest penalty being 10 months' imprisonment, or a fine of up to HK$70,000; that those 74 cases included three involving products claimed to be "health food" but found in fact to contain controlled drug ingredients; and that over the past three years (2021 to 2023) the Department of Health reviewed 128 365 advertisements and issued 1 344 warning letters [Note 30]. The Department of Health's own appeal to consumers also states that members of the public are strongly urged not to buy products of unknown or dubious composition, and not to take products of unknown origin, since their safety, quality and efficacy are not assured [Note 30].

On supplements, the Department of Health's Elderly Health Service has Chinese text of its own. Its page on vitamins and health states that the best source of vitamins is food; that vitamin supplements cannot replace food and can only make up for nutritional shortfalls caused by illness; that to date no study has established that taking vitamin supplements in normal health makes the body stronger; and that the doses in most supplements on the market far exceed the recommended intakes, so that long-term use may produce side effects — excessive vitamin A supplementation causing liver toxicity, for instance, and high-dose vitamin C supplements possibly causing gastrointestinal upset and increasing the chance of kidney and urinary stones — so a doctor or dietitian should be consulted first [Note 31].

⚠️ A substantive divergence between the Chinese and English versions, reported here as found, with no splitting of the difference. The corresponding passage in the English version of the same page reads "there is little evidence to show that vitamin intakes higher than the suggested level can further promote health in healthy individuals" — which is about intakes above the suggested level. The Chinese is about taking vitamin supplements as such. These are two different statements, and the Chinese one is the broader. This article follows the project's convention of quoting the Chinese original, and will not read the English meaning into the Chinese quotation.

The closing passage of the same service's guide to cancer prevention and antioxidants (February 2026, Chinese only) also states that there is at present no definitive data establishing that any single nutrient or substance can wholly prevent cancer, so that beyond taking in more foods rich in antioxidants, the everyday diet must also follow a healthy pattern low in fat, sugar and salt and high in fibre, in order to reduce the risk of cancer [Note 31].

⚠️ If you are about to buy an "anti-ageing" supplement online, hold two things together: (i) claims of that kind are not among the six classes named in the Fourth Schedule; and (ii) of the Department of Health's 74 convictions for unregistered pharmaceutical products over the three years 2021 to 2023, three involved products claimed to be "health food" but found to contain controlled drug ingredients. "Marketed as health-promoting" and "reviewed and registered" are two different things.

#Hong Kong against elsewhere, and how the Government itself puts it

The Hong Kong Government does publish international comparisons, but what it publishes is a comparison against one named list, not a world ranking.

Table 14 of Hong Kong Population Projections 2022–2046, "Expectation of Life at Birth in Hong Kong and Selected Economies, 1991 to 2021", column for 2021 (the table lists 14 economies in all):

Do the arithmetic (calculated here): on the 2021 column of that table, Hong Kong men at 83.2 years are 1.7 years above Japan's 81.5 and 9.7 years above the United States' 73.5; Hong Kong women at 87.9 years are 0.3 years above Japan's 87.6. These are subtractions performed here on the table's figures, not gaps published by the department.

Table 15 of the same document projects Hong Kong forward: men from 83.2 years in 2021 (an actual figure) to 86.4 years in 2046; women from 87.9 to 91.8.

⚠️ Why this article uses a named list rather than a ranking. Neither of the two departmental publications cited above (Hong Kong Life Tables 2016–2046, 70 pages, and Hong Kong Population Projections 2022–2046, 74 pages) makes any world-ranking statement at all — what they produce is the named-economy list above. World-ranking statements appear in a different class of document: Policy Addresses and bureau blogs (see the opening of this article). The distinction is worth noticing: the publications that produce the figures make only named comparisons, and the ranking language appears in policy documents.

⚠️ If you are writing, presenting or arguing and need to cite Hong Kong life expectancy: a named-list comparison (the table above) is far safer than a ranking, because a ranking depends on which body compiled it, for which year, and against which list of places — and the department's own two publications use precisely a named list rather than a ranking. If you must use a ranking, cite its source with it (which Policy Address, which year), because the Government's own wording already differed between 2022 and 2023.

#What to do next

There is only one sourced set of actions this article can point to — and it is not any author's opinion here, it is text the Centre for Health Protection has published.

Page 1 of the Centre for Health Protection's Non-Communicable Diseases Watch of February 2015 carries a checklist titled the keys to ageing well, introduced by the sentence that healthy ageing begins with healthy behaviours in youth; its thirteen items are reproduced in full at [Note 32]: do not smoke; be physically active; eat a healthy and balanced diet; maintain a desirable body weight and waist circumference; avoid alcohol; get enough sleep; keep the mind agile; manage stress with a positive attitude; keep close ties with family, friends and society; prevent accidents and injury; maintain oral health; use medicines safely; and obtain appropriate preventive medical and health services.

⚠️ That checklist comes from the Centre for Health Protection's Non-Communicable Diseases Watch of February 2015, and is reproduced here in full; the source itself makes no claim that the list is exhaustive. Not being on this list does not mean something is unimportant, and lists published by different bodies differ.

As for how much exercise specifically, this article will only point back to the WHO recommendation quoted above: at least 150–300 minutes of moderate-intensity aerobic activity a week, and muscle strengthening on two days or more a week; for those aged 65 and above, add activity emphasising balance and strength on three days or more a week. This article will not issue any personalised exercise prescription, will not specify any intensity zone, and will not recommend any supplement or drug. Anyone with cardiovascular risk or a long-term condition, or moving from no exercise at all to exercising, should speak to a doctor before starting.

And before paying for any "longevity" product or service, this article suggests putting it into that gap and asking once: is this product changing a prognostic marker, or something that a randomised controlled trial has already shown changes the outcome? So far, for every item this article has looked at — VO2max, grip strength, biological age, rapamycin, micro-dose lithium, NDGA — the answer is the former.

⚠️ One last limit, and it is this article's own: it has not gone through the interventional trial evidence for each of the behaviours above one by one. The pattern described here — strong association, thin interventional evidence — is stated for VO2max and grip strength; it cannot be transferred automatically to smoking cessation, diet, sleep or the rest, some of which (smoking cessation, for one) have interventional evidence of a quite different character that this article has not examined. The difference is this: nobody charges you for these behaviours, they carry no prescription drug's list of side effects, and they carry no unregistered product's ingredient risk. Where the strength of the evidence is comparable, the difference in cost and risk is everything.

#Frequently asked questions

In the end, does exercise really extend life?

In observational studies the association is both strong and consistent — Mandsager 2018's cohort of 122,007, Kokkinos 2022's of 750,302 and Leong 2015's PURE study of 139,691 all point the same way. But the one large long-term randomised controlled trial with mortality as its endpoint (Generation 100, 1,567 people, five years) produced no difference between groups, and every one of its confidence intervals crosses 1. Note that the control group in that trial was not "no exercise" but "follow the national guidelines" — so it cannot answer "exercise against no exercise", only "supervised structured exercise against following the guidelines on your own".

So should I go and have my VO2max or grip strength measured?

This article makes no recommendation on individual tests. What can be said is that the PURE paper itself describes grip strength as a "simple, inexpensive risk-stratifying method" — stratification is not diagnosis, and it is not the same as changing that number changing the outcome. The same study also found no significant association between grip strength and incident diabetes, fall injury or fracture.

Should I still be doing Zone 2?

This article does not write exercise prescriptions. What can be said is that the conclusion of the narrative review by Storoschuk et al. in Sports Medicine in 2025 is that current evidence does not support Zone 2 training as the optimal intensity for improving mitochondrial or fatty acid oxidative capacity — that is, the conclusion is confined to those two endpoints — and that higher intensities should be prioritised where training volume is lower; while equally, that review is narrative rather than systematic, one of its authors holds equity in an exercise-related US company, and no trial compares Zone 2 with other intensities on mortality.

Did Bryan Johnson add low-dose lithium and NDGA in 2026?

This article found nothing to support it. On a recount as at 3 August 2026, a word-by-word count of his public protocol site (a single page for the whole site) returned 0 occurrences each for rapamycin, lithium and NDGA (with Acarbose, Tadalafil and sauna used as positive controls on the same page, returning 2, 2 and 29 respectively, showing the count itself works). But his other domain answered the crawl with HTTP 429 and could not be examined — so this is "not written on the public page", not "established as not being taken".

Rapamycin extends mouse lifespan by 23–26% — can people take it?

Every registered sirolimus product in Hong Kong is a Part 1, Schedule 1 and Schedule 3 poison and a prescription-only medicine, meaning a registered doctor's prescription is required. The 23–26% is median lifespan, in mice, at three times the dose (Miller 2014). On the human side, the only randomised controlled trial of any size (PEARL, 48 weeks, 114 completing) had a null primary endpoint, its registered sponsor AgelessRx is itself a company that publicly sells rapamycin (with UCLA as a collaborator), and the formulation used reached only about a third of the blood concentration of the commercial product.

Does micro-dose lithium protect the brain?

The association in the Danish drinking-water study (Kessing 2017, 73,731 cases against 733,653 controls) is nonlinear: relative to 2.0–5.0 µg/L, the incidence rate ratio above 15.0 µg/L is 0.83, but the 5.1–10.0 µg/L band is 1.22 (that is, 22% higher). The 2025 Nature paper is a mouse and human tissue study with no human trial. In Hong Kong, lithium carbonate is a prescription medicine — this article's search reached only that name and does not cover other lithium salts such as lithium orotate.

Are there safety problems with NDGA supplements?

There is a signal, and it needs saying. NDGA comes mainly from chaparral (Larrea tridentata); the US National Library of Medicine's LiverTox records more than two dozen published cases of liver injury since 1990, some severe, some leading to emergency liver transplantation and some apparently leading to cirrhosis, with recurrence on reexposure and a likelihood score of B; the same entry also states that chaparral preparations "have not been shown to be effective in any medical condition". Note that LiverTox is describing the plant extract, not purified NDGA at a stated dose. NDGA is not on the Hong Kong register of pharmaceutical products (five spellings all returned zero, with metformin as a positive control confirming the query works) — but that does not establish that it cannot be bought in Hong Kong.

Is a biological age test worth having?

That is the subject of a separate article and is not opened up here. This article records one related datum only: the epigenetic age sub-study of the PEARL trial (24 people) states in the original that "we saw no meaningful significant changes between groups".

Do people in Hong Kong live the longest in the world?

The Hong Kong Government has said so, but different departments in different years have used formulations of very different scope. Paragraph 100 of the Chief Executive's 2022 Policy Address says Hong Kong has the longest life expectancy in the world; paragraph 113 of the 2023 Policy Address changes it to among the longest life expectancies in the world; and the Census and Statistics Department, which produces the figures, writes in Statistics and Us (November 2025) that it is among the longest in the developed economies. The three range from first in the world to near the top among developed economies. If you are going to quote one, the department's is the narrowest and the closest to the figures it publishes itself.

#Notes: the original texts

[Note 1] Kokkinos P et al., J Am Coll Cardiol 2022;80(6):598-609, conclusion: "Being unfit carried a greater risk than any of the cardiac risk factors examined."

[Note 2] Storoschuk KL et al., Sports Med 2025;55(7):1611-1624: "These recommendations largely stem from observational data of elite endurance athletes who engage in large volumes of Zone 2 training and possess high mitochondrial and fatty acid oxidative capacity. However, we challenge the broad endorsement of Zone 2 training for members of the general public, as it contradicts substantial evidence supporting the use of high-intensity exercise for improving mitochondrial capacity and cardiometabolic health." And "We conclude that current evidence does not support Zone 2 training as the optimal intensity for improving mitochondrial or fatty acid oxidative capacity. Further, evidence suggests prioritizing higher exercise intensities (> Zone 2) is critical to maximize cardiometabolic health benefits, particularly in the context of lower training volumes."

[Note 3] LiverTox, "Chaparral" entry, US National Library of Medicine: "Chaparral extracts have been linked to several cases of clinically apparent liver injury, some of which have led to acute liver failure and need for emergency liver transplantation." "Liver injury attributable to chaparral was first reported in 1990 and subsequently more than two dozen cases of clinically apparent liver injury attributed to chaparral have been published mostly from the United States, Canada and Australia. The time to onset varied from 3 weeks to several years, but was usually within 3 to 12 weeks of starting daily ingestion or increasing the daily dose. … Several reported cases have been severe and some have led to emergency liver transplantation, others appeared to result in cirrhosis. … Despite the several reports of liver injury caused by chaparral, over-the-counter products with chaparral are still available commercially and on the internet. For unclear reasons, there have been no cases of liver injury clearly implicating chaparral published since 2005." "Likelihood score: B (likely but now rare cause of clinically apparent liver injury)." "Recurrence after reexposure has occurred." And two consecutive sentences from the Background section: "Chaparral is also claimed to retard aging and aid in wellbeing. However, chaparral preparations have not been shown to be effective in any medical condition." The same entry also writes that NDGA "has been used as a food additive in low concentrations".

[Note 4] Census and Statistics Department interactive data table 115-01011, Chinese symbol note:

Because the mortality rate in 2022 was unusually high during the 2019 coronavirus disease epidemic, the expectation of life at birth for 2022 should be interpreted with care. (our translation from the Chinese original)

Chinese original:

「由於2022年死亡率在2019冠狀病毒病疫情期間異常地高，請小心詮釋2022年的出生時平均預期壽命。」

[Note 5] Census and Statistics Department, Hong Kong Life Tables 2016–2046 (2022 edition), paragraph 2.4 and paragraph 2.6:

2.4 The expectation of life at birth for men rose from 67.8 years in 1971 to 83.2 years in 2021, an increase of 15.5 years over 50 years. The expectation of life at birth for women rose from 75.3 years in 1971 to 87.9 years in 2021, an increase of 12.6 years over 50 years. / 2.6 The gap between the expectation of life at birth for men and for women narrowed from 7.5 years in 1971 to 5.7 years in 2019. That gap fluctuated slightly over 2020 to 2022 during the 2019 coronavirus disease epidemic. It is projected that the gap will remain at around 5.4 years to 5.5 years from 2023 to 2046. (our translation from the Chinese original)

Chinese original:

「2.4 男性的出生時平均預期壽命，由1971 年的67.8 年增加至2021 年的83.2 年，在50 年間增加了15.5 年。女性的出生時平均預期壽命，由1971 年的75.3 年增加至2021 年的87.9 年，在50 年間增加了12.6 年。」 「2.6 男性及女性出生時平均預期壽命的差距，由1971 年的7.5 年收窄至2019 年的5.7 年。該差距在2019 冠狀病毒病疫情期間的2020 年至2022 年略有波動。按推算該差距在2023 年至2046 年會保持在5.4 年至5.5 年左右。」

[Note 6] Chief Executive's 2022 Policy Address, paragraph 100: "100. Hong Kong has the longest life expectancy in the world." The same report, paragraph 146, also states that it is the longest-living place in the world. 2023 Policy Address, paragraph 113: "113. Hong Kong has among the longest life expectancies in the world", in a paragraph noting that the birth rate remains low, that the average number of children per couple fell to a new low of 0.9 in 2022, and that the share of the population aged 65 and above will climb from 20% to nearly a third within the next ten years. Census and Statistics Department, Statistics and Us (November 2025):

People in Hong Kong live very long lives. In 2024 the expectation of life at birth was 83 years for men and 88 years for women, both about 7 years more than 30 years ago. The expectation of life at birth for men and women in Hong Kong is among the longest in the developed economies. (our translation from the Chinese original)

Chinese original:

「香港人十分長壽。2024年，男性的出生時平均預期壽命是83年，女性則是88年，兩者較30年前均增加了約7年。香港男性及女性的出生時平均預期壽命，在已發展的經濟體而言是最長之一。」

[Note 7] Census and Statistics Department, Hong Kong Population Projections 2022–2046, paragraph 2.4:

2.4 The population is expected to continue ageing. … Excluding foreign domestic helpers, the elderly population will rise from 1.45 million in 2021 (20.5% of the total population) by 1.29 million to 2.74 million in 2046 (36.0%). Compared with the increase of about 0.8 million over the preceding 25 years (1996 to 2021), the pace of increase in the elderly population is clearly accelerating. (our translation from the Chinese original)

Chinese original:

「2.4 未來人口預期持續高齡化。⋯⋯撇除外籍家庭傭工，長者人口由2021 年的145 萬（佔總人口的20.5%）上升129 萬至2046 年的274 萬（36.0%）。相較之前25 年（1996 年至2021 年）約80 萬的升幅，未來長者人口上升的速度明顯加快。」

[Note 8] Centre for Health Protection, Department of Health, Non-Communicable Diseases Watch, February 2015, "Ageing well", page 3:

Research shows that people at higher health risk (risk defined by smoking, a higher body mass index, and physical inactivity in middle and later adult life) develop disability earlier in later life and have more of it. Conversely, adopting a healthier lifestyle not only extends life but also lengthens the time spent in health. Compared with the high-risk group, the age at which disability appears in the low-risk group can be postponed by more than five years. / In Hong Kong, a thematic household survey conducted between October 2011 and January 2012 found that nearly three quarters (73.7%) of people aged 65 and above had a chronic disease diagnosed by a Western medicine practitioner, including hypertension (46.0%), diabetes (20.0%), heart disease (9.8%), cancer (3.5%) and stroke (3.2%). (our translation from the Chinese original)

Chinese original:

「研究顯示，較高健康風險的人士（風險由吸煙、較高的體重指數、中年和成年後期缺乏體能活動等界定）在晚年較早出現殘疾，而且患的殘疾也較多。相反地，採取較健康的生活模式不僅延長壽命，亦令享有健康的時間更長。與高風險組別作比較，低風險組別出現殘疾的年齡可延遲五年以上」 「在香港，於二零一一年十月至二零一二年一月進行的主題性住戶統計調查發現，接近四分之三（73.7%）年齡在65歲及以上的人士患有經西醫診斷的慢性疾病，包括高血壓（46.0%）、糖尿病（20.0%）、心臟病（9.8%）、癌症（3.5%）及中風（3.2%）。」

[Note 9] Mandsager K et al., JAMA Netw Open 2018;1(6):e183605: "Risk-adjusted all-cause mortality was inversely proportional to cardiorespiratory fitness and was lowest in elite performers (elite vs low: adjusted hazard ratio [HR], 0.20; 95% CI, 0.16-0.24; P < .001; elite vs high: adjusted HR, 0.77; 95% CI, 0.63-0.95; P = .02)." And "The increase in all-cause mortality associated with reduced cardiorespiratory fitness (low vs elite: adjusted HR, 5.04; 95% CI, 4.10-6.20; P < .001; below average vs above average: adjusted HR, 1.41; 95% CI, 1.34-1.49; P < .001) was comparable to or greater than traditional clinical risk factors (coronary artery disease: adjusted HR, 1.29; 95% CI, 1.24-1.35; P < .001; smoking: adjusted HR, 1.41; 95% CI, 1.36-1.46; P < .001; diabetes: adjusted HR, 1.40; 95% CI, 1.34-1.46; P < .001)."

[Note 10] Kodama S et al., JAMA 2009;301(19):2024-35: "Participants were categorized as low CRF (< 7.9 METs), intermediate CRF (7.9-10.8 METs), or high CRF (> or = 10.9 METs)."

[Note 11] Leong DP et al., Lancet 2015;386(9990):266-73: "Grip strength was inversely associated with all-cause mortality (hazard ratio per 5 kg reduction in grip strength 1.16, 95% CI 1.13-1.20; p<0.0001), cardiovascular mortality (1.17, 1.11-1.24; p<0.0001), non-cardiovascular mortality (1.17, 1.12-1.21; p<0.0001), myocardial infarction (1.07, 1.02-1.11; p=0.002), and stroke (1.09, 1.05-1.15; p<0.0001). Grip strength was a stronger predictor of all-cause and cardiovascular mortality than systolic blood pressure." "We found no significant association between grip strength and incident diabetes, risk of hospital admission for pneumonia or COPD, injury from fall, or fracture." And the conclusion: "Further research is needed to identify determinants of muscular strength and to test whether improvement in strength reduces mortality and cardiovascular disease."

[Note 12] Momma H et al., Br J Sports Med 2022;56(13):755-763: "J-shaped associations with the maximum risk reduction (approximately 10-20%) at approximately 30-60 min/week of muscle-strengthening activities were found for all-cause mortality, CVD and total cancer" and the conclusion: "the influence of a higher volume of muscle-strengthening activities on all-cause mortality, CVD and total cancer is unclear when considering the observed J-shaped associations"

[Note 13] Blair SN et al., JAMA 1995;273(14):1093-8: "The highest age-adjusted all-cause death rate was observed in men who were unfit at both examinations (122.0/10,000 man-years); the lowest death rate was in men who were physically fit at both examinations (39.6/10,000 man-years). Men who improved from unfit to fit between the first and subsequent examinations had an age-adjusted death rate of 67.7/10,000 man-years. This is a reduction in mortality risk of 44% (95% confidence interval, 25% to 59%) relative to men who remained unfit at both examinations."

[Note 14] Stensvold D et al., BMJ 2020;371:m3485. Sampling frame: "General population of older adults in Trondheim, Norway". Results section: "…ty of 4.7%), an absolute risk reduction of 1.7 percentage points was observed after HIIT (hazard ratio 0.63, 95% confidence interval 0.33 to 1.20) and an absolute increased risk of 1.2 percentage points after MICT (1.24, 0.73 to 2.10). When HIIT was compared with MICT as reference group an absolute risk reduction of 2.9 percentage points was observed (0.51, 0.25 to 1.02) for all cause mortality." Conclusion: "This study suggests that combined MICT and HIIT has no effect on all cause mortality compared with recommended physical activity levels. However, we observed a lower all cause mortality trend after HIIT compared with controls and MICT." The control group's actual intensity: "Control participants chose to perform more of their physical activity as HIIT than the physical activity undertaken by participants in the MICT group. This meant that the controls achieved an exercise dose at an intensity between the MICT and HIIT groups."

[Note 15] Declaration of interests for Storoschuk et al. 2025: "Martin J. Gibala is an advisor to and holds equity in Longevity League, Ltd., a US-based company whose services in part relate to exercise."

[Note 16] Centre for Health Protection, Department of Health, "Physical Activity" health information page (page dated 15 April 2025):

For adults aged 18 and above (including those living with a chronic condition or disability), the World Health Organization recommends: / They should undertake regular physical activity. / They should do at least 150–300 minutes of moderate-intensity aerobic physical activity a week; or at least 75–150 minutes of vigorous-intensity aerobic physical activity; or at least an equivalent amount of a combination of moderate- and vigorous-intensity activity, for substantial health benefits. / They should also do muscle-strengthening activities at moderate or greater intensity targeting all major muscle groups on two days or more each week, which brings additional health benefits. / For older people aged 65 and above (including those living with a chronic condition or disability), they should also, on 3 days or more each week, do varied physical activity at moderate or greater intensity emphasising balance and strength training, as part of their weekly physical activity, to enhance functional capacity and prevent falls. / They may increase moderate-intensity aerobic physical activity to more than 300 minutes a week; or do more than 150 minutes of vigorous-intensity aerobic physical activity; or an equivalent amount of a combination of moderate- and vigorous-intensity activity, for further health benefits. / The Behavioural Risk Factor Survey of 2023 indicates that 14.8% of people aged 18 and above are insufficiently physically active, 16.0% among women and 13.4% among men. (our translation from the Chinese original)

Chinese original:

「對18歲及以上（包括患有慢性病或殘疾）的成年人來說，世界衞生組織建議：」「他們應定期進行體能活動。」「他們應每星期進行最少150–300分鐘中等強度的帶氧體能活動；或最少75–150分鐘劇烈強度的帶氧體能活動；或最少相等於混合中等和劇烈強度活動模式的時間，以獲得顯著健康裨益。」「他們還應每星期有兩天或以上，進行中等或更高強度針對所有主要肌肉群的強化肌肉活動，這能帶來額外健康裨益。」「適用於65歲及以上（包括患有慢性病或殘疾）的長者，他們亦應每星期有3天或以上，進行多種著重平衡和力量訓練的中等或更高強度體能活動，作為每星期體能活動的一部分，以提升身體功能和預防跌倒。」「他們可以將每星期中等強度的帶氧體能活動增加到300分鐘以上；或進行150分鐘以上劇烈強度的帶氧體能活動；或相等於混合中等和劇烈強度活動模式的時間，以獲得更多健康裨益。」及「2023年度健康行為調查指出，有14.8%的18歲或以上人士體能活動量不足，女士中有16.0%，男士中有13.4%。」

[Note 17] protocol.bryanjohnson.com (retrieved 2 August 2026), foot of page: "The protocol presented on this website is based on scientific research, ranging from mouse studies to meta-analyses of randomized controlled trials and international clinical practice guidelines. These have been carefully reviewed for their unique relevance to my personal situation. The protocol encompasses a mix of on-label, off-label, and unlicensed therapies, as well as research-use-only tests. Some of these tests and therapies are still under scientific investigation and have not yet received on-label licensing for specific health conditions. All tests and therapies, regardless of their licensing status, carry risks. These risks have been assessed for my personal use by a specialized team of clinicians and scientists. This protocol represents an experimental clinical research project."

[Note 18] Bryan Johnson, "I stopped taking rapamycin" (publication date shown on the page 01.25.2025): "Despite the immense potential from pre-clinical trials, my team and I came to the conclusion that the benefits of lifelong dosing of Rapamycin do not justify the hefty side-effects (intermittent skin/soft tissue infections, lipid abnormalities, glucose elevations, and increased resting heart rate)."

[Note 19] Sehgal R et al., bioRxiv 2024 (DOI 10.1101/2024.10.22.619522), abstract: "Here we curate TranslAGE-Response, a harmonized database of 51 public and private longitudinal interventional studies and calculate a consistent set of 16 prominent epigenetic clocks for each study, along with 95 other DNAm biomarkers that help explain changes in each clock." The statement carried on every page: "which was not certified by peer review".

[Note 20] protocol.bryanjohnson.com, biological age section: "These tests are experimental and intended solely for research purposes. They should not replace or supplement any clinical tests recommended by licensed medical professionals."

[Note 21] Drug Office, Department of Health, "Search Drug Database" for pharmaceutical products registered in Hong Kong: "RAPAMYCIN IS THE SYNONYM / STANDARD NAME OF : sirolimus" "Record: 1 to 3 of 3" and that database's definition: "Prescription Only Medicines are medicines which must only be purchased with a prescription in a pharmacy." (The database's Chinese page states itself that it is available in English only.)

[Note 22] Moel M et al., Aging (Albany NY) 2025;17(4):908-936: "Adverse and serious adverse events were similar across all groups. Visceral adiposity did not change significantly (ηp2 = 0.001, p = 0.942), and changes in blood biomarkers remained within normal ranges. Lean tissue mass (ηp2 = 0.202, p = 0.013) and self-reported pain (ηp2 = 0.168, p = 0.015) improved significantly for women using 10 mg rapamycin. Self-reported emotional well-being (ηp2 = 0.108, p = 0.023) and general health (ηp2 = 0.166, p = 0.004) also improved for those using 5 mg rapamycin. No other significant effects were observed." Conclusion: "Future work will evaluate benefits of a broader range of rapamycin doses on healthspan metrics for longevity, and will aim to more comprehensively establish efficacy." The epigenetic sub-study: "Within the epigenetic testing results, we saw no meaningful significant changes between groups."

[Note 23] The same, declaration of interests: "GH, VL, AN, MM, SM, AI, and SZ are employees and shareholders of AgelessRx."

[Note 24] The same, full text: "It was subsequently discovered that compounded rapamycin did indeed have approximately ⅓ the concentration in blood after 24 hrs relative to commercial [40]. As such, while rapamycin doses are listed at the advertised compounded dose, it should be noted that equivalent effective doses for compounded forms are approximately 66% less."

[Note 25] Roark KM, Iffland PH II, Front Aging 2025;6:1628187: "Further, the long-term effects and safety of chronic mTOR inhibition in healthy humans and whether rapamycin can truly "slow" human aging or prevent age-related diseases without unacceptable side effects is unknown." And "As rapamycin gains popularity for its anti-aging potential, online longevity clinics have emerged offering access to the drug with minimal medical oversight. This semi-regulated availability raises ethical concerns regarding patient safety, misinformation, and the potential for serious harm."

[Note 26] Kessing LV et al., JAMA Psychiatry 2017;74(10):1005-1010: "A nonlinear association was observed. Compared with individuals exposed to 2.0 to 5.0 µg/L, the incidence rate ratio (IRR) of dementia was decreased in those exposed to more than 15.0 µg/L (IRR, 0.83; 95% CI, 0.81-0.85; P < .001) and 10.1 to 15.0 µg/L (IRR, 0.98; 95% CI, 0.96-1.01; P = .17) and increased with 5.1 to 10.0 µg/L (IRR, 1.22; 95% CI, 1.19-1.25; P < .001)."

[Note 27] Aron L et al., Nature 2025;645(8081):712-721: "Li replacement with amyloid-evading salts is a potential approach to the prevention and treatment of AD."

[Note 28] Strong R et al., Aging Cell 2008;7(5):641-50: "Comparison of the proportion of live mice at the age of 90% mortality was used as a surrogate for measurement of maximum lifespan; neither NDGA (p=0.12) nor aspirin (p=0.16) had a significant effect in this test." Strong R et al., Aging Cell 2016;15(5):872-84: "The effects of NDGA were dose dependent and male specific but without an effect on maximal lifespan."

[Note 29] US Code of Federal Regulations, 21 CFR § 189.165: "(a) Nordihydroguaiaretic acid is the chemical 4,4′-(2,3-dimethyltetramethylene) dipyrocatechol, C18H22O4. It occurs naturally in the resinous exudates of certain plants. The commercial product, which is synthesized, has been used as an antioxidant in foods." "(b) Food containing any added NDGA is deemed to be adulterated in violation of the act based upon an order published in the Federal Register of April 11, 1968 (33 FR 5619)."

[Note 30] Written reply of the Secretary for Health, Professor Lo Chung-mau, to Legislative Council question twenty-two on the regulation of health food (5 June 2024):

"Health food" covers many kinds of product of differing composition and character, and internationally there is no consistent definition or regulation of "health food"; similar products may go under many different names, such as dietary supplements, nutritional supplements and natural health foods. / As for products meeting the definition of "pharmaceutical product" under the Pharmacy and Poisons Ordinance and of "proprietary Chinese medicine" under the Chinese Medicine Ordinance, they must satisfy the requirements of the relevant ordinances as to safety, quality and efficacy, and must be registered before they may be sold or supplied in Hong Kong. / The Government also regulates, through the Undesirable Medical Advertisements Ordinance, the labelling and promotion of products making medical or health claims (including products that are not "pharmaceutical products" or "proprietary Chinese medicines"), so that members of the public do not place undue trust in medical or health claims and self-treat, thereby delaying seeking medical attention. / At this stage the Health Bureau considers that, from a risk perspective, the existing framework regulates the safety of such products to an appropriate degree, and that strengthening the relevant public education and provision of information is more suitable than directly regulating the health claims of "health food". / Note: the six claims are (i) preventing, eliminating or treating breast lumps, (ii) regulating the function of the genito-urinary system, (iii) regulating the endocrine system, (iv) regulating the sugar or glucose content of the body, (v) regulating blood pressure and (vi) regulating blood lipids or cholesterol. / Over the same period the Department of Health handled 74 cases resulting in conviction for unregistered pharmaceutical products, the heaviest penalty being 10 months' imprisonment, or a fine of up to HK$70,000. … Those 74 convictions for unregistered pharmaceutical products included three cases involving products claimed to be "health food" but found in fact to contain controlled drug ingredients. … Over the past three years, up to 2023, the Department of Health reviewed 128 365 advertisements and issued 1 344 warning letters. / Members of the public are strongly urged not to buy products of unknown or dubious composition, and not to take products of unknown origin, since their safety, quality and efficacy are not assured. (our translation from the Chinese original)

Chinese original:

「「保健食品」種類繁多，成分各異，性質多樣，國際間對「保健食品」並沒有一致的定義和監管，類似的產品可包含很多不同名目，例如膳食補充劑、營養補充劑及天然健康食品等。」「至於符合《藥劑業及毒藥條例》定義的「藥劑製品」及《中醫藥條例》定義的「中成藥」必須按相關條例要求符合安全、品質及成效，並經註冊後才可在香港作出銷售及供應。」「政府亦透過《不良廣告（醫藥）條例》監管作出醫療或保健聲稱的產品（包括並非「藥劑製品」或「中成藥」的產品）的標籤及宣傳，以免市民過分相信醫療或保健聲稱，自行治理，導致延誤求醫。」「現階段醫務衞生局認為從風險角度而言，現有框架對有關產品的安全性已有適度規管，而加強相關的公眾宣傳教育和資訊提供會比直接規管「保健食品」的健康聲稱較合適。」「註：該六項聲稱包括（一）預防、消除或治療乳房腫塊、（二）調節生殖泌尿系統的機能、（三）調節內分泌系統、（四）調節體內糖分或葡萄糖、（五）調節血壓及（六）調節血脂或膽固醇。」「同期，衞生署共處理了74宗涉及未經註冊藥劑製品而被定罪的個案，最高刑罰為被判監禁10個月，或罰款最高港幣70,000元。⋯⋯上述74宗未經註冊藥劑製品定罪個案中，包括三宗聲稱為「保健食品」的產品而實際發現含受管制藥物成分的個案。⋯⋯過去三年（二○二一至二○二三年），衞生署審視了128 365則廣告，並發出1 344封警告信。」「強烈呼籲市民切勿購買成分不明或可疑的產品，亦切勿服用來歷不明的產品，因其安全、素質及效能均未獲保證」

[Note 31] Department of Health, Elderly Health Service, "Vitamins and health":

The best source of vitamins is food, because it is most readily absorbed and used by the body, and because food contains, besides vitamins, other substances beneficial to the body such as minerals, fibre, antioxidants and phytochemicals; so vitamin supplements cannot replace food, and can only make up for nutritional shortfalls caused by illness. To date no study has established that taking vitamin supplements in normal health makes the body stronger. / The vitamin supplements now on the market are of many kinds, with differing formulations and doses, and are not necessarily suitable for everyone; moreover the doses of the vitamins mostly far exceed the recommended intakes, and long-term use may produce side effects — excessive vitamin A supplementation causing liver toxicity, for instance, and high-dose vitamin C supplements possibly causing gastrointestinal upset and increasing the chance of kidney and urinary stones. Anyone intending to take a supplement should guard against excess harming their health, and should therefore consult a doctor or dietitian first. (our translation from the Chinese original)

Chinese original:

「維生素的最佳來源是食物，因為它最易為人體吸收和利用，而且食物除含有維生素外，亦含其他對身體有益的物質如礦物質丶纖維素丶抗氧化物質丶植物素等，所以維生素補充劑並不能代替食物，而只能彌補因疾病所引致營養攝取的不足。到目前為止尚未有研究證實於健康正常情況下服用維生素補充劑會令身體更強健。」 「現時市面售賣的維生素補充劑種類繁多，配方及劑量都各有不同，未必人人適用，而且維生素的劑量大多遠超過建議的攝取量，長期服用有可能會產生副作用，例如過量服用維生素A補充劑會引致肝中毒，高劑量的維生素C補充劑可能會令腸胃不適，以及增加患腎石及尿道石的機會。若擬服用補充劑，要提防出現過量的情況，影響健康，所以應先諮詢醫生或營養師。」

(The character 「丶」 in the original is as it appears on the source page, reproduced verbatim.) The closing passage of the same service's guide to cancer prevention and antioxidants (February 2026):

At present there is no definitive data establishing that any single nutrient or substance can wholly prevent cancer, so the everyday diet, besides taking in more foods rich in antioxidants, must also follow a healthy dietary pattern low in fat, sugar and salt and high in fibre, in order to reduce the risk of developing cancer. (our translation from the Chinese original)

Chinese original:

「目前沒有確切數據證明任何單一營養素或物質能完全預防癌症，所以日常飲食，除了應多攝取含豐富抗氧化物的食物外，還須遵循低脂、低糖、低鹽、高纖維的健康飲食模式，以減低罹癌的風險。」

[Note 32] Centre for Health Protection, Department of Health, Non-Communicable Diseases Watch, February 2015, page 1, the checklist of keys to ageing well (the first line is the introductory sentence):

Healthy ageing begins with healthy behaviours in youth. / Do not smoke / Be physically active / Eat a healthy and balanced diet / Maintain a desirable body weight and waist circumference / Avoid alcohol / Get enough sleep / Keep the mind agile / Manage stress with a positive attitude / Keep close ties with family, friends and society / Prevent accidents and injury / Maintain oral health / Use medicines safely / Obtain appropriate preventive medical and health services (our translation from the Chinese original)

Chinese original:

「康健樂頤年始於年輕時的健康行為。」「不要吸煙」「積極參與體能活動」「健康及均衡飲食」「維持理想的體重及腰圍」「避免飲酒」「充足睡眠」「鍛煉腦筋靈活」「積極態度，管理壓力」「與家人、朋友和社會保持密切關係」「預防意外和受傷」「保持口腔健康」「安全使用藥物」「獲取適當的預防性醫療和健康服務」

#What this article does not state

Each of the following was looked for and no citable source was found, so no statement is made — with a record of what was searched, so that readers can judge for themselves.

• This sentence: "raising VO2max or grip strength extends life" — this article does not state it. This is its central limitation. All the large data available are observational associations; the only long-term randomised controlled trial with mortality as its endpoint (Generation 100) produced no difference between groups; and the PURE paper states in writing that "whether improvement in strength reduces mortality" still needs testing.
• These two sentences: "Zone 2 training extends life" and "Zone 2 training does not extend life" — this article says neither. No randomised controlled trial was found comparing Zone 2 with other intensities on mortality, lifespan or hard disease endpoints; ClinicalTrials.gov was searched (3 August 2026) and "zone 2" AND exercise returned 15 studies, most of them surgery on zone two of the flexor tendons of the hand, with none having mortality or lifespan as an endpoint. Storoschuk 2025's argument rests on mitochondrial capacity and cardiorespiratory fitness, not on mortality.
• Three statements in circulation with no first-hand source, not carried here (reasons in the Bryan Johnson section above): USD 2,000,000 spent a year; "about 80% of the effect can be reproduced for under USD 100 a month", attributed to Johnson himself; and a named doctor "still taking rapamycin".
• The exact year Bryan Johnson stopped taking rapamycin — the post says only "On September 28th", with no year; the post is dated 25 January 2025. From the post's own sequence (the preprint it goes on to mention was posted in October 2024) it can be inferred to be 2024, but that is an inference, not a date the post states, so this article presents it as an inference and does not cite it as fact.
• This sentence: "a preprint shows rapamycin accelerating ageing on 16 epigenetic clocks" — this article makes no directional statement. The preprint has been identified through item 5 of the post's own Sources list (Sehgal R et al., bioRxiv, DOI 10.1101/2024.10.22.619522, PMID 39484592), and both the "16 clocks" figure and rapamycin's inclusion among the interventions analysed are written in the preprint itself. But the directional conclusion "accelerates ageing" appears only in Johnson's restatement and not in the preprint's own text; this article did not obtain the individual effect values and therefore restates no direction. The preprint has also not been peer reviewed. The other PMC article popularly cited for this claim (PMC12226543) contains, counted word by word, "epigenetic" 0 times and "clock" 0 times, and has nothing to do with the subject.
• This sentence: "sirolimus is not approved in Hong Kong for anti-ageing use" — the Hong Kong pharmaceutical products registration database does not publish approved indications, only legal classification, sale requirement and registration certificate holder. This article therefore makes no such statement, and says only that this source publishes no approved indications.
• The content of blueprint.bryanjohnson.com — on 2 August 2026 the root, the product list and a direct curl request to that domain all returned HTTP 429 (local_rate_limited) for about 15 minutes. Every statement here that the three compounds are not on his public protocol is therefore confined in scope to protocol.bryanjohnson.com.
• A human longevity or healthspan randomised controlled trial for micro-dose lithium or NDGA — there is none. The clinical trial registries were searched (ClinicalTrials.gov, 3 August 2026, see the table above): lithium orotate one study (Alzheimer's disease, not yet recruiting), NDGA nine (all oncology), none with longevity or healthspan as an endpoint; the positive control metformin AND aging returned 55 and the negative control 0. On the literature side, three lines were searched: drinking water and dementia, endogenous lithium deficiency and Alzheimer's disease, and case reports of lithium orotate supplement toxicity.
• NDGA's status in US food law — established, see the NDGA section above: 21 CFR 189.165 sits within a part titled Substances Prohibited From Use in Human Food, and since 1968 any food containing added NDGA is deemed adulterated. Its position under the US dietary supplement regime, and under Hong Kong food law, were not checked and no statement is made about either.
• Prices for rapamycin, lithium or NDGA in Hong Kong — no published Hong Kong price was found.
• Cardiorespiratory fitness or grip strength cohort data for Hong Kong or East Asian populations — Mandsager, Kokkinos, Blair and Generation 100, all cited here, are all Western populations; PURE covers China but its abstract does not report a separate estimate for China. This article therefore makes no statement about the size of these associations in Hong Kong people.
• Absolute survival figures, per-site denominators and confidence intervals from the NIA Interventions Testing Program mouse studies — only the abstracts were obtained this time, not the full texts, so these are not listed.
• Recent actual published percentages for the population aged 65 and above from the Census and Statistics Department — the 20% and 34% cited here are projected values from Hong Kong Population Projections 2022–2046, not actual annual figures; the actual figures could not be located in the department's published material this time.
• Harms of GLP-1 drugs in people who are not obese — this crawl did not cover the subject at all, and this article makes no statement about it.
• Enrolment in the Chronic Disease Co-Care Scheme and the extension date for hepatitis B screening — not covered by this crawl and not carried here.
• Sleep and salt intake figures from the Population Health Survey 2020-22 — not covered by this crawl. The local physical activity figure used here comes from a different and more recent survey (the Behavioural Risk Factor Survey of 2023), and the two cannot be mixed.
• Prices for GrimAge and other biological age tests at private Hong Kong laboratories — a figure of "about HK$2,000–5,000" circulates; this crawl cannot support it and this article does not carry it. That subject belongs to a separate article.

#Sources

• Census and Statistics Department interactive data table 115-01011, vital events / expectation of life at birth (men and women, 2021–2025; "p" = provisional figure; "‡" = the 2022 epidemic note): https://www.censtatd.gov.hk/tc/web_table.html?id=115-01011 and English version https://www.censtatd.gov.hk/en/web_table.html?id=115-01011 (table data release date 30 June 2026; last revised 9 July 2026; retrieved 2 August 2026)
• Census and Statistics Department, Hong Kong Life Tables 2016–2046, 2022 edition (paragraph 2.4 on the 1971→2021 increase; paragraph 2.6 on the gap between the sexes and the 2023–2046 projection; table 24 and the discrepancy between its 2022 figures and the interactive data table): https://www.censtatd.gov.hk/en/data/stat_report/product/B1120016/att/B1120016092023XXXXB01.pdf (the document itself: published August 2023; retrieved 2 August 2026)
• Census and Statistics Department, Hong Kong Population Projections 2022–2046 (table 1 on the share aged 65 and above, the elderly dependency ratio and the median age; paragraph 2.4 on the elderly population excluding foreign domestic helpers; table 14 comparing fourteen economies in 2021; table 15 on the Hong Kong projection): https://www.censtatd.gov.hk/en/data/stat_report/product/B1120015/att/B1120015092023XXXXB01.pdf (the document itself: published August 2023; retrieved 2 August 2026)
• Census and Statistics Department, Statistics and Us ("among the longest in the developed economies"; 83 years for men and 88 for women in 2024): https://www.censtatd.gov.hk/tc/page_235.html and English version https://www.censtatd.gov.hk/en/page_235.html (the page itself: November 2025; retrieved 3 August 2026)
• Chief Executive's 2022 Policy Address, paragraph 100 ("Hong Kong has the longest life expectancy in the world"): https://www.policyaddress.gov.hk/2022/tc/p100.html and English version https://www.policyaddress.gov.hk/2022/en/p100.html (the English original: "Hong Kong has the longest life expectancy in the world."; retrieved 3 August 2026)
• Chief Executive's 2022 Policy Address, paragraph 146 (the longest-living place in the world): https://www.policyaddress.gov.hk/2022/tc/p146.html (retrieved 3 August 2026)
• Chief Executive's 2023 Policy Address, paragraph 113 ("Hong Kong has among the longest life expectancies in the world"): https://www.policyaddress.gov.hk/2023/tc/p113.html and English version https://www.policyaddress.gov.hk/2023/en/p113.html (the English original: "Hong Kong has among the longest life expectancies in the world"; retrieved 3 August 2026)
• Census and Statistics Department interactive data table 110-01004, dependency ratios and median age of the population (definition of the elderly dependency ratio and its values for 2021–2025; median age 48.4 in 2025): https://www.censtatd.gov.hk/tc/web_table.html?id=110-01004 (retrieved 2 August 2026)
• Centre for Health Protection, Department of Health, "Physical Activity" health information page (the Chinese original of the World Health Organization's physical activity recommendations for adults and for people aged 65 and above; 14.8% insufficiently active in the Behavioural Risk Factor Survey of 2023; the two typographical errors in the English version): https://www.chp.gov.hk/tc/healthtopics/content/100200/8804.html and English version https://www.chp.gov.hk/en/healthtopics/content/100200/8804.html (page dated 15 April 2025; retrieved 2 August 2026)
• Centre for Health Protection, Department of Health, Non-Communicable Diseases Watch, February 2015, "Ageing well" (the full checklist of keys to ageing well; healthspan and the postponement of disability by more than five years; 73.7% comorbidity among people aged 65 and above in the thematic household survey of 2011–12): https://www.chp.gov.hk/files/pdf/ncd_watch_feb2015_chin.pdf and English version https://www.chp.gov.hk/files/pdf/ncd_watch_feb2015.pdf (publication date February 2015; retrieved 2 August 2026)
• Department of Health, Elderly Health Service, "Vitamins and health" (the Chinese original statements on vitamin supplements; the substantive divergence between the Chinese and English versions): https://www.elderly.gov.hk/tc_chi/healthy_ageing/healthy_diet/eating_nutrition/vitamins_and_health.html and English version https://www.elderly.gov.hk/english/healthy_ageing/healthy_diet/eating_nutrition/vitamins_and_health.html (the page carries no revision date; retrieved 2 August 2026)
• Department of Health, Elderly Health Service, guide to cancer prevention and antioxidants (no single nutrient or substance can wholly prevent cancer): <https://www.elderly.gov.hk/tc_chi/education_and_media_resources/files/防癌攻略 - 抗氧化篇 (2026年2月).pdf> (article dated February 2026, Chinese only; retrieved 2 August 2026)
• Legislative Council question twenty-two: regulation of health food, written reply of the Secretary for Health, Professor Lo Chung-mau (the definition of health food; registration required for pharmaceutical products and proprietary Chinese medicines; the list of six classes of claim regulated under the Undesirable Medical Advertisements Ordinance; 74 convictions in 2021–2023, the heaviest penalty, 128 365 advertisements and 1 344 warning letters; the Department of Health's appeal to consumers): https://www.info.gov.hk/gia/general/202406/05/P2024060500372.htm and English version https://www.info.gov.hk/gia/general/202406/05/P2024060500373.htm (5 June 2024; retrieved 2 August 2026)
• Drug Office, Department of Health, "Search Drug Database" for pharmaceutical products registered in Hong Kong (the legal classification and sale requirement of the three registered sirolimus/rapamycin products; seven for lithium carbonate; zero results for all five spellings of NDGA, nordihydroguaiaretic acid and masoprocol, with metformin as a positive control): https://www.drugoffice.gov.hk/eps/drug/productSearchOneField (the database itself: Last Updated: 31-Jul-2026; that search is available in English only; retrieved 2 August 2026)
• Mandsager K, Harb S, Cremer P, Phelan D, Nissen SE, Jaber W. Association of Cardiorespiratory Fitness With Long-term Mortality Among Adults Undergoing Exercise Treadmill Testing. JAMA Netw Open 2018;1(6):e183605 (122,007 people; median follow-up 8.4 years; 13,637 deaths; elite against low HR 0.20; the comparison against coronary artery disease, smoking and diabetes) (DOI 10.1001/jamanetworkopen.2018.3605; PMID 30646252)
• Kokkinos P, et al. Cardiorespiratory Fitness and Mortality Risk Across the Spectra of Age, Race, and Sex. J Am Coll Cardiol 2022;80(6):598-609 (750,302 US veterans; median follow-up 10.2 years; 174,807 deaths; lowest risk at about 14.0 METs; least fit 20% HR 4.09) (DOI 10.1016/j.jacc.2022.05.031; PMID 35926933)
• Kodama S, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA 2009;301(19):2024-35 (33 studies, 102,980 people; RR 0.87 per 1 MET; low against high fitness RR 1.70; the three MET bands) (DOI 10.1001/jama.2009.681; PMID 19454641)
• Blair SN, et al. Changes in physical fitness and all-cause mortality. A prospective study of healthy and unhealthy men. JAMA 1995;273(14):1093-8 (9,777 men; age-adjusted death rates of 122.0, 67.7 and 39.6 per myriad man-years across the three groups; a 44% lower risk in those who went from unfit to fit) (PMID 7707596)
• Stensvold D, et al. Effect of exercise training for five years on all cause mortality in older adults—the Generation 100 study: randomised controlled trial. BMJ 2020;371:m3485 (1,567 people; HIIT 400 / MICT 387 / control 780; control-group mortality 4.7%; the three hazard ratios and their confidence intervals; the control group's actual exercise intensity) (DOI 10.1136/bmj.m3485; ClinicalTrials.gov NCT01666340; PMID 33028588)
• Leong DP, et al. Prognostic value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study. Lancet 2015;386(9990):266-73 (17 countries, 139,691 people; median follow-up 4.0 years; 3,379 deaths; the hazard ratios per 5 kg lower; no association with diabetes, falls or fracture; the authors' own statement that whether raising strength lowers mortality remains to be tested) (DOI 10.1016/S0140-6736(14)62000-6; PMID 25982160)
• Momma H, Kawakami R, Honda T, Sawada SS. Muscle-strengthening activities are associated with lower risk and mortality in major non-communicable diseases: a systematic review and meta-analysis of cohort studies. Br J Sports Med 2022;56(13):755-763 (16 studies; 10–17% lower risk; J-shaped associations with the maximum reduction at approximately 30–60 minutes a week) (DOI 10.1136/bjsports-2021-105061; PMID 35228201)
• Storoschuk KL, Moran-MacDonald A, Gibala MJ, Gurd BJ. Much Ado About Zone 2: A Narrative Review Assessing the Efficacy of Zone 2 Training for Improving Mitochondrial Capacity and Cardiorespiratory Fitness in the General Population. Sports Med 2025;55(7):1611-1624 (current evidence does not support Zone 2 as the optimal intensity; higher intensities to be prioritised where training volume is lower; the authors' declaration of interests) (DOI 10.1007/s40279-025-02261-y; PMID 40560504)
• Bull FC, et al. World Health Organization 2020 guidelines on physical activity and sedentary behaviour. Br J Sports Med 2020;54(24):1451-1462 (the academic source of the recommendation the Centre for Health Protection quotes) (DOI 10.1136/bjsports-2020-102955; PMID 33239350)
• Laukkanen T, Khan H, Zaccardi F, Laukkanen JA. Association between sauna bathing and fatal cardiovascular and all-cause mortality events. JAMA Intern Med 2015;175(4):542-8 (2,315 Finnish men; median follow-up 20.7 years; sudden cardiac death HR 0.37 against a reference group using a sauna once a week; HR 0.48 for sessions longer than 19 minutes; no significant association with all-cause mortality) (DOI 10.1001/jamainternmed.2014.8187; PMID 25705824)
• Harrison DE, et al. Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature 2009;460(7253):392-5 (feeding started at 600 days; 14% in females and 9% in males measured at the age of 90% mortality) (DOI 10.1038/nature08221; PMID 19587680)
• Miller RA, et al. Rapamycin-mediated lifespan increase in mice is dose and sex dependent and metabolically distinct from dietary restriction. Aging Cell 2014;13(3):468-77 (median lifespan up 23% in males and 26% in females at three times the dose) (DOI 10.1111/acel.12194; PMID 24341993)
• Moel M, et al. Influence of rapamycin on safety and healthspan metrics after one year: PEARL trial results. Aging (Albany NY) 2025;17(4):908-936 (48 weeks; 114 completing, 40/35/39; null primary endpoint of visceral adiposity; the compounded formulation at about a third of the blood concentration; the declaration of AgelessRx employees and shareholders; the epigenetic sub-study of 24 with no significant change; the itemised serious and non-serious adverse events) (DOI 10.18632/aging.206235; PMID 40188830; PMCID PMC12074816)
• Roark KM, Iffland PH II. Rapamycin for longevity: the pros, the cons, and future perspectives. Front Aging 2025;6:1628187 (the long-term effects of chronic mTOR inhibition in healthy humans being unknown; online longevity clinics offering the drug with minimal medical oversight) (DOI 10.3389/fragi.2025.1628187; PMID 40620657)
• Kessing LV, et al. Association of Lithium in Drinking Water With the Incidence of Dementia. JAMA Psychiatry 2017;74(10):1005-1010 (73,731 against 733,653; a nonlinear association; IRR 1.22 at 5.1–10.0 µg/L) (DOI 10.1001/jamapsychiatry.2017.2362; PMID 28832877)
• Aron L, et al. Lithium deficiency and the onset of Alzheimer's disease. Nature 2025;645(8081):712-721 (a mouse model plus human tissue; lithium orotate used in mice; the authors' own framing as a "potential approach") (DOI 10.1038/s41586-025-09335-x; PMID 40770094)
• McKnight RF, et al. Lithium toxicity profile: a systematic review and meta-analysis. Lancet 2012;379(9817):721-8 (385 studies; the figures for renal function, thyroid, parathyroid and weight; scope confined to therapeutic doses in patients with mood disorders) (DOI 10.1016/S0140-6736(11)61516-X; PMID 22265699)
• Pauzé DK, Brooks DE. Lithium toxicity from an Internet dietary supplement. J Med Toxicol 2007;3(2):61-2 (an 18-year-old woman, 18 capsules, each of 120 mg lithium orotate; serum lithium 0.31 mEq/L; the elemental lithium conversion of 3.83 mg per 100 mg against 18.8 mg per 100 mg) (DOI 10.1007/BF03160910; PMID 18072162)
• Strong R, et al. Nordihydroguaiaretic acid and aspirin increase lifespan of genetically heterogeneous male mice. Aging Cell 2008;7(5):641-50 (male mice p=0.0006; no significant effect on the surrogate measure of maximum lifespan, p=0.12) (DOI 10.1111/j.1474-9726.2008.00414.x; PMID 18631321)
• Strong R, et al. Longer lifespan in male mice treated with a weakly estrogenic agonist, an antioxidant, an α-glucosidase inhibitor or a Nrf2-inducer. Aging Cell 2016;15(5):872-84 (the original dose and doses three times higher and lower all reproduced it; dose dependent and male specific; no effect on maximum lifespan) (DOI 10.1111/acel.12496; PMID 27312235)
• LiverTox: Clinical and Research Information on Drug-Induced Liver Injury, "Chaparral" entry, US National Library of Medicine: https://www.ncbi.nlm.nih.gov/books/NBK548355/ (more than two dozen published cases of liver injury since 1990; acute liver failure and emergency transplantation; recurrence after reexposure; likelihood score B; "Chaparral is also claimed to retard aging" and the sentence immediately refuting it; NDGA as its principal constituent and "has been used as a food additive in low concentrations") (retrieved 3 August 2026)
• US Code of Federal Regulations, 21 CFR § 189.165 "Nordihydroguaiaretic acid (NDGA)", within subpart C of part 189, Substances Prohibited From Use in Human Food (the order published on 11 April 1968, 33 FR 5619; food containing added NDGA deemed adulterated): https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-189/subpart-C/section-189.165 (retrieved 3 August 2026)
• Sehgal R, Borrus D, Kasamato J, et al. DNAm aging biomarkers are responsive: Insights from 51 longevity interventional studies in humans. bioRxiv, 2024 (the preprint indicated by item 5 of the Sources list on Johnson's post; 51 interventional studies, 16 epigenetic clocks; not peer reviewed) (DOI 10.1101/2024.10.22.619522; PMID 39484592; PMCID PMC11526957)
• ClinicalTrials.gov registration record NCT04488601 (the PEARL trial; Lead Sponsor AgelessRx, agency class INDUSTRY; Collaborator the University of California, Los Angeles; Responsible Party Sponsor; primary endpoint visceral adiposity measured by DXA): https://clinicaltrials.gov/study/NCT04488601 (retrieved 3 August 2026)
• ClinicalTrials.gov search results (lithium orotate 1; nordihydroguaiaretic OR NDGA OR masoprocol 9; "zone 2" AND exercise 15; positive control metformin AND aging 55; negative control 0): https://clinicaltrials.gov/ (query date 3 August 2026)
• AgelessRx company product page, "Rapamycin Rx" (marked "Starting at $65 / mo" and "Compounded and generic options"): https://agelessrx.com/rapamycin/ (retrieved 3 August 2026)
• Bryan Johnson's public protocol page (the five prescription medicines in the Rx column; the whole-page statement of limits; the statement of limits on biological age testing; the statements in the sauna section): https://protocol.bryanjohnson.com/ (retrieved 2 August 2026)
• Bryan Johnson, "I stopped taking rapamycin" (the reasons for stopping and the four side effects; "On September 28th" with no year given): https://blueprint.bryanjohnson.com/blogs/news/i-stopped-taking-rapamycin (publication date shown on the page 25 January 2025; retrieved 2 August 2026)
• The list prices of six retail products on the Blueprint online store home page (archived snapshot): https://web.archive.org/web/20260515145600/https://blueprint.bryanjohnson.com/ (snapshot time 15 May 2026; retrieved 2 August 2026)

This article was written from the sources listed above; information date 2 August 2026. It is general health information, does not constitute medical advice, and does not recommend any drug, supplement, test, course of treatment or service provider. Individual diagnosis, medication, dosage and exercise plans are for a doctor to decide.

---

#Further reading

• Bryan Johnson's public protocol page: protocol.bryanjohnson.com
• Bryan Johnson's original post on stopping rapamycin: blueprint.bryanjohnson.com/blogs/news/i-stopped-taking-rapamycin
• Department of Health, Population Health Survey 2020–22 report (Part II): chp.gov.hk
• Chronic Disease Co-Care Scheme (CDCC): primaryhealthcare.gov.hk/cdcc
• On this platform: Biological age tests
• On this platform: The evidence behind longevity science