Review on Bryan Johnson case

Epistemologic status: I spent about 4 hours to analyze Bryan Johnson case, mostly the main website which I read fully (and several other his articles). I'm fairly convinced in my conclusions and open for discussion. Probably some of my criticism is already addressed somewhere - in that case I'd like to hear about that and will edit the review accordingly.
Bryan Johnson is a great person who motivates many people to be better. In 45 years he has excellent physical form and good health which both seem to improve quite fast. However some people tend to overestimate his achievements.

So here are my initial thoughts on the website https://blueprint.bryanjohnson.co/:
1) He doesn't substantiate anything with links to scientific articles; there are almost no such links. Almost all of his "(ref)" lead to a page on his website with general words about how to choose biomarkers, not to actual references to relevant scientific articles.
How to review his statements in that case if I don't know which scientific papers he relies on? It's possible but much harder.
He writes about "+ 1,000 Peer-reviewed References" but they are not published on the website.

2) Many of his biomarkers are bad.
He wakes up in the middle of the night with severe headaches: "Severe headache symptoms causing to wake in the night, on acetazolamide 250 mg since Jan 25th, 2022". I don't think that's quite normal for 45 year old person.
In these links:

• https://blueprint.bryanjohnson.co/#heart-health
  
• https://blueprint.bryanjohnson.co/#brain-health
  
• https://blueprint.bryanjohnson.co/#ear-health
  

you can see that many of his biomarkers are much older than his 45 year old age, some correspond to 70+ years
Here is another list of his health problems: https://blueprint.bryanjohnson.co/#notable-challenges
RIJV: 70-80% stenosis
LIJV: 100% stenosis
Missing left transverse sinus on MRV; cerebral venous thrombosis?
Can't make quick assessment of how bad it is.

3) He seems to put the main emphasis on good physical fitness and its corresponding biomarkers (VO2max, max leg press, max bench press, etc.). His numbers are very good and motivational, but are not really surprising (for someone of his age, motivation and budget). That approximately what you would expect from a very motivated person (even in 45 years) who can afford to invest much into own physical form.

4) He also emphasizes the reduction of the arithmetic mean of the 6 epigenetic clocks.
He is 45 years old. In April he had an epigenetic age of 47, in November 42.5.
4.1) A separate big topic, of course, is how well epigenetic clocks predict mortality and life expectancy, and which epigenetic clocks predict it best. Let's assume that we believe in their relevance.
4.2) Decreasing epigenetic age from 47 to chrono-age 45 is one thing, it should be easier. Low base effect (bad start). So if we speak of real rejuvenation, it's mostly from 45 to 42.5.
4.3) Another question - what is the average deviation for matching chrono and epi clocks? The suspicion is that even though he sort of rejuvenated his epi-clock, he rather went from +0.5 sigma to -0.5 sigma of average epi-age for his current age, or something like that. That's cool, but maybe not a real rejuvenation.
4.4) not all of his epi clocks had significant rejuvenation. For example, PCPhenoAge was 40.9, became 40.6. AgeHorvath IEAA decreased by 2.6 years (while PCHorvath1 and PCHorvath2 decreased by 6.3 and 5.5 years, respectively). We need to figure out why such a difference between these epigenetic clocks here? Why some of them change easily and others are difficult to change? Which of epigenetic clocks are better at predicting mortality - those easily changed or those hard to change?

5) He writes that he has 6.9% fat, and that his body temperature has dropped by 3 degrees Fahrenheit. Okay. But is that really optimal for life expectancy? That needs to be studied separately, but as far as I've analyzed before, 7% fat is not something that's better than say 12% or 15% in terms of life expectancy. Same issue with too much physical activity. There are studies that this may not be optimal for life expectancy.

6) About his interventions.
(6.1) In the food section ("Step 1: Meal Prep") I didn't find anything special that wasn't discussed and practiced by many life extensionists. Except that he eats chlorella to get spermidine.
(6.2) In the supplements section ("Step 2: Supplements"), everything is pretty standard, too. Between all the garlic, vitamin E and BroccoMax stuff, the only things worth noting are:

• testosterone 2mg patch 6x weekly
  
• lithium 1mg
  
• Rapamycin 13mg, bi-weekly (Rx)
  

Though, why not.
Overall, there's nothing special about his supplements or his food (or his workouts... and even his described accomplishments, while generally cool, aren't anything extraordinary for someone with his motivation and budget).
(6.3) and by the way, no references to scientific articles or at least to any guidelines of anything. Of course it is clear that he makes decisions about food and supplements on the basis of tests... but this should also be based on some scientific basis, on scientific papers which we'd like to see. Compare for example to Australian dietary guidelines which have >1100 references to scientific papers.

(7) https://blueprint.bryanjohnson.co/#current-results
He writes that his 26 biomarkers are perfect (Ctrl+F "Results: Optimal Clinical Outcome Range")
while also ~120 "younger than Chronological Age".
(7.1.1) As for 120 biomarkers, how much are they younger? a month, a year? How many biomarkers did he measure to get these 120 which happened to be younger than his chronological age? 200? 300? We don't know.
(7.1.2) Many of these 120 biomarkers are likely "younger" (better) for innate causes (like most likely "cerebral volume", "whole brain volume", "white matter volume" are), many are simply related to his good shape, and 11 biomarkers, for example, are just associated with the number of wrinkles on the face, etc. (every rich person can afford to have younger face nowadays, nothing new). We could make a short list of the most important biomarkers out of these 120, but see (7.1.1), it is useless, he does not give the detailed information on how much younger they are.
(7.2) About the ideal 26 biomarkers... Well, he's on testosterone, so two of those 26 ideal biomarkers are predictably Testosterone и Free Testosterone Index. Another his "ideal biomarker" is 6.9% body fat. Not sure at all that 6.9% body fat is better for life expectancy compared to 12% or 15%. Two of "ideal biomarkers" are about grip strength. Well ok, if he trains for it, then grip strength will be better than it normally is (though 134lbs =~60kg is not extraordinarly much for 45 years old person). It is possible (if necessary) to select a short list of the most important ideal biomarkers from these 26, I'm just not sure that anyone needs it after all what has been said.

Overall, nothing extraordinary as of 30.01.2023. Still it's an interesting and motivational case. Maybe further large improvements in his health and biomarkers would make me to change my mind.

===
More on (4.4):
In paper https://pubmed.ncbi.nlm.nih.gov/33211845/ Horvath himself writes that the "first generation" HorvathAA and HannumAA epichases are not really useful. Table 2 of the paper shows that Horvath has no data that HorvathAA and HannumAA can predict all-cause mortality. PhenoAgeAA is slightly better, though statistical significance has not yet been shown. But GrimAgeAA predicted all-cause mortality well. The HRs with CIs for all-cause mortality for one standard unit (z-score) increase are: for HorvathAA, one model gives 0.97 (0.70, 1.36), second 1.03 (0.74, 1.44); for HannumAA, 0.89 (0.64, 1.24) and 0.92 (0.67, 1.28); for PhenoAgeAA, 1.26 (0.92, 1.74) and 1.13 (0.81, 1.57) (finally, at least increase in epigenetic age leads to increase in all-cause mortality in average); for GrimAgeAA, 2.05 (1.45, 2.90) and 1.91 (1.23, 2.96) (immensely better than previous clocks).

Horvath himself writes that "HorvathAA and HannumAA were not predictive of health", "The results of this study confirm Horvath and Raj's (5) previous observation that the first-generation clocks are not sensitive predictors of age-related decline in clinical health measures. Few of the associations were significant, and even among those that were, the observed relationships were in the opposite direction to what one would hypothesize"[1].
By the way, Bryan wrote something about GrimAge, not on the main page, but in articles. For example, in January 2022 he promised that "we will continue to publish updates on these 6 clock results, as well as GrimAge, approximately quarterly"[2].
Interestingly, in October 2021 Bryan wrote that his GrimAge epigenetic age is 36: "My chronological age is 44, measured biological age is 36" (it's about GrimAge, see the first picture on the link)[3]. It would be interesting to see the dynamics of GrimAge, where he seems to have great start.
Correlations of epigenetic age with chronological age can be seen in the Supplementary to Horvath's article https://pubmed.ncbi.nlm.nih.gov/33211845/ FigS1. You can see that a deviation of plus or minus 5-10 years from the epigenetic age corresponding on average to your chrono-age is widespread. GrimAge (and PhenoAge) have slightly smaller deviations.