Horvath Epigenetic Age: A Weak Spot, But I Have A Plan (14-Test Analysis)
Вставка
- Опубліковано 5 чер 2024
- Join us on Patreon! / michaellustgartenphd
Discount Links:
Epigenetic Testing: trudiagnostic.com/?irclickid=...
Use Code: CONQUERAGING
NAD+ Quantification: www.jinfiniti.com/intracellul...
Use Code: ConquerAging At Checkout
Oral Microbiome: www.bristlehealth.com/?ref=mi...
Enter Code: ConquerAging
At-Home Metabolomics: www.iollo.com?ref=michael-lustgarten
Use Code: CONQUERAGING At Checkout
At-Home Blood Testing (SiPhox Health): getquantify.io/mlustgarten
Green Tea: www.ochaandco.com/?ref=conque...
Diet Tracking: shareasale.com/r.cfm?b=139013...
If you'd like to support the channel, you can do that with the website, Buy Me A Coffee: www.buymeacoffee.com/mlhnrca
Conquer Aging Or Die Trying Merch! my-store-d4e7df.creator-sprin...
------------------------------------------------------------------------------------------------------------------------------------------------------------------
Papers referenced in the video:
Underlying features of epigenetic aging clocks in vivo and in vitro
onlinelibrary.wiley.com/doi/1...
Thumbnail image via:
Back to the future: Epigenetic clock plasticity towards healthy aging
pubmed.ncbi.nlm.nih.gov/29337... - Наука та технологія
having a good plan is always good. 😎
i really admire your work!
Thanks @monnoo8221!
This adventure sure has twists and turns. Science is fun. Wild!
This information you've provided today I believe confirms the more or less consensus in the longevity community at lower protein, lower total fat, and certainly lower saturated fat is what is beneficial.
I think for the experiment's sake of the next test, you should lower the protein to lower than 95 g. Maybe like 70 to 80 g or you could even try the .8 g per kilo of body weight.
Thanks Steve-it's only a correlation, so I can't say if it's causation. Also note that the diet isn't as low protein as advertised for longevity, with ~100g/d = 1.6g/kg BW/d, which is ~2x the RDA. Similarly, fat intake (including the gut microbiome's contribution) isn't low-fat, with >40% of calories coming from fat. But the SFA data is in line with general recommendations...
That's tricky for protein, as I could risk losing muscle mass. During my 20 lb weight loss, there have been very minor strength losses (i.e. overhead press, 90-5 reps at a higher body weight, now 90-3 reps), and cutting protein intake could accelerate that. In contrast, cutting fat intake (but not too much) is on the current experimental list...
Intuitively, this makes sense. Lower animal protein, lower sat fat should only have positive impact.
We'll see, you might be right!
@@conqueragingordietrying1797 Look up the new Nurses cohort paper on protein and healthy aging. Animal protein is bad for chronic disease, plant protein is fantastic.
Very nice video, very objective and correlates with many findings from other researchers, like protein intake for example. 👍
Thanks @aljosarojac8575!
tbh i am even more intererested on what worsens the Horvath clock , tks for sharing the bad results as well as the good
Thanks @barrie888, and always on the data, whether good or bad!
Thanks for the video. This is one of the most unusual and controversial biological age results compared with your previous results from other biological age method. I thought the biological age measured from the DNA methylation should be less fluctuated compared with other blood biomarkers method. But it seems to be varied a lot and might be too "sensitive". It also does not aligned with all your good results (all biomarkers optimized) which I supposed to be rather reliable. I am also not very convinced that higher protein intake would result in shorter DNA methylation. I am sure there is something wrong and you would figure it out finally. Look forward to seeing your next video. Great work!
Thanks @wflai98.Horvath is a molecular clock, which could be an early warning system of physiological big-picture changes (physical, cognitive function), but also, when considering that they're farthest away from those outcomes, would be expected to be more variable relative to levels of cells (WBC, RBC) or proteins (albumin, hsCRP, etc).
That I can have youthful biomarkers despite a terrible Horvath (for now) suggests that I have other compensatory youthful mechanisms, but that may not always be the case-I'd like to optimize Horvath, too, to minimize any weak spots before they're all that I see.
I wonder if eugenol in cloves has anything to do with the association. It is a histamine receptor antagonist, and I guess you could always take antihistamines to test that pathway. I don't remember if you have any allergies or have done extensive intolerance testing.
Ah, I do have seasonal allergies in the summer.
Ideally, I'd get vaccinated against tree pollen to potentially fix that once-and-for-all, rather than trying to block it with antihistamines.
I think cloves have both the highest anti-glycation capacity and have the highest antioxidant measurements in some studies. Eugenol, in one study, reduces serum testosterone levels and sperm viability in adult Wistar rats. Not sure what it does in humans. I used to chew on some cloves, but I think the Rat study may have caused me to end that practice.
You can get cronometer data from apple health export and then run a small script to format it however you want saving as excel or csv. Presumably it would save you a bunch of time manually entering into a spreadsheet?
Ah, I have no coding experience-the key would be lining up the exported data for each individual food to be in its own column (i.e. collated).
Although daily food intake can be exported, Cronometer doesn't collate it, which makes it easier (in my case) to manually enter it into a Spreadsheet. In contrast Cronometer collates macros and micros for a given period.
Are some of the previous clocks reported from earlier generations of the Horvath clock, or have those results been recalculated with the latest generation clock?
I'm not sure-this should be the most recent Horvath iteration, but I'll have to check
Check for the just released pre-print paper "ComputAgeBench: Epigenetic Aging Clocks Benchmark" it does includes horvath v1/v2 as well
Nice, thanks @utyep23985uhX
would be cool to see if its not just total protein (which it might well be) but if its more about specific amino acids that mess up the numbers at higher intakes (and maybe if there is even ones that do the opposite, like maybe glycine or NAC 😁)
I supplemented with 5g of BCAAs/d for Test #15, so we'll see how that goes...
@@conqueragingordietrying1797 May I ask why?
I've heard from Morgan Levine that Horvath clock (and methylation clocks in general) can vary by 5-10 years within a single day. If that's true then do you try to control your testing timing?
Yep, I test at the same time of day, same fasting window before testing, same hydration status.
The only way to account for its test-to-test variability is to measure more often, so I'm aiming for 10 tests in 2024, vs 8 in '23.
As we have talked about before, any methylation based epigenetic test is positively correlation with methyl donor intake, especially niacin..Can you analyze that with your past data?
I did-Horvath is not significantly correlated with dietary TMG intake, nor other potential methyl donors (proline betaine, tryptophan betaine, trigonelline, not the sum of these methylated metabolites. B12, B6, and folate are also not significantly correlated, so there's more to the story than methylation...
Coconut butter is basically driving that sfa and total fat, right? Just curious, what inspired eating coconut butter in general?
Likely yes-it was positively correlated with HDL in my data, but more recently, it looks like a high Vitamin K1 intake is driving that association.
@@conqueragingordietrying1797 Yes K1 seems to act primarily in the liver. The associations with K1 and HDL are stronger in those deficient in K1, and if you were to push K1 high enough you'd probably see a flattening.
@@jamesgilmore8192 Yep, potentially getting there at 25-30x the RDA (2500 - 3000 mcg/d)
@@conqueragingordietrying1797 Besides vitamin K1, is there anything else that increases HDL?
Mike -- what type of sample are you sending for horvath? Is it a blood spot, saliva, blood draw, tasso?
Blood spot
@@conqueragingordietrying1797 So its probably responding mostly to RBCs. Those last 90 days or so. I wonder if the epigenome is being mostly changed in marrow or while in circulation? I'm not sure if any research has been done on that. If say the epigenome is being modified while in the marrow then you are probably correlating on the wrong timeframe. Perhaps you could redo the correlations by reweighting past food intake in a more sophisticated way. I wonder if this is why you are having trouble with the optimisation of the epigenome in general.
While I deeply respect your diiligent analysis and restricting variables to only dietary modifications, I speculate that you will eventually include pharmaceuticals before (beyond your thyroid med). Maybe I'm way off and you'd get plasma dilution before that ;-) , but if you were to BioHack Aging with pharma, would your first choice be a rapa? Please continue your excellence, sir.
Thanks @stonebridge7710, I already do, including low-dose nicotinic acid (for NAD) and Vitamin D, but not at the magnitude for others in this space
I'm open to rapa, especially because my Candida IgG is higher than I like. This could be on the list sooner vs later...
@@conqueragingordietrying1797I’m surprised you haven’t thrown in LDN and Mythlene blue into your mix yet lol! Those two actually saved my life and cured my Parathesia
Whats your assessment on finasteride effect on skin aging?
There is some evidence that DHT may promote fibroblast degeneration and thus decreasing systemic DHT may help slowing down this process
I haven't studied it, but I'm interested in potentially slowing skin aging via sulforaphane (from broccoli sprouts in my case):
pubmed.ncbi.nlm.nih.gov/34271100/
@@conqueragingordietrying1797 Astaxanthin and curcumin are other options for Nrf2 activators.
A 10 year variability over just two years seems odd. I'm wondering about the reliability of the Horvath test.
I agree, but it can be overcome with lots of testing, which is the plan
Thank you, negative results are just as informative as positive, but rarely are people forthcoming with that. Of course, this could just be a random outlier.
I think you should try a 5-day, low protein, low calorie, near-fast. I think you can still total the same amount of everything averaged over a testing period, just eating a little more than usual, on a few other days. We were never meant to always have abundance. Our bodies are designed to switch gears and drop IGF-1 and then start cleaning up cells, and deleting crummy cells, sacrificing them for protein to continue functions. We don't want to stay there a long time. When food is restored, IGF-1 returns and your body builds up new replacement cells.
The protein you eat looks very good. I don't think it is aging you, I think you just want a break from it for a few days.
First time I did this was amazing. I cut my calories to 1/4-1/5 normal, and less than 9% of that being protein. I ate a lot of green beans, soup, butternut squash, summer squash, bok choy, and some potato. Daily, I had one date and one walnut at the same time. I drank hibiscus tea, water, favored, stevia sweetened (actual stevia not the cut commercial stuff in the groceries), and water. But I suspect you can continue to only eat the things that are normally in your diet, just a lot less of them, and temporarily omit some of the high protein foods. I definitely ate more salt than normal. Needed it to make the soups work.
This is worth it just for the brain boost.
Some advocates suggest doing fasting mimicking monthly or every few months. I don't think that is optimal. I think once a year to once every two years.
This is not a weight loss strategy. This is just a natural way to clean up cells. I took fisetin. You take that already, just eating all those strawberries. I took my regular supplements too and counted in those calories in the spreadsheet.
Thanks @ChessMasterNate. I try to eat close to maintenance body weight intake, as dramatic cuts below that (even for 5d) have been historically bad for being able to regulate calorie intake and body weight afterwards.
are you still able to build muscle with that amount of protein? it's crazy to see your protein goals, since most pig out on it.
Yep, I have a decent amount of muscle: instagram.com/p/C6pC9DjRGJX/
@@conqueragingordietrying1797 I'm plant-based and eat around 85 - 105g of protein daily.
5'4. 127 lbs and have a decent amount of muscle for my stature.
I only wonder about increasing my protein to see if that affects recovery/muscle knots.
You could try an AKG supplement. AKG is essential for TET activation and there is some evidence in a poorly designed study that it impacted age in a cheap but nonstandard clock.
I'm not sold on AKG-that video is on the to-do list...
Can you explain why your horvath biological age is so much higher than your biological age from other calculators like the excel spreadsheet one you use?
Horvath is at the molecular level (i.e epigenetic changes), which could take time to mess up systemic, further away from molecular biomarkers (metabolites, cells, proteins, etc).
Hopefully Horvath results isn't a future prediction of what my other data will become, and I'm going to collect more data (at least 8 tests in 2024) to try to figure it out asap.
Horvath was trained on lots of different tissues and cell types and it it likely picking up "global" effects of the epigenome. It's all cause mortality isn't as strong as some of the other clocks that have been more extensively constrained and optimised.
@@jamesgilmore8192 its association to ACM is rather weak. The whole methylation clock approach is a big myth or at lest what I would call a weak theory.
IMHO the change of classical physiological parameters, over a sufficiently wide vector, is much more indicative.
@@jamesgilmore8192 do you know which is the best one? dunedinpace?
Do you purchase TruAge Complete or TruAge Pace each time? Does it send you back one number, or many numbers? I'm interested, but it's definitely a steep cost. Thanks for the discount links and good luck on the next round! I love following your progress
Thanks Gail! I'm confident I'll figure it out, but ha, I hope it doesn't take 40 tests to get there.
TruAge Complete (many numbers), which includes Horvath, DunedinPACE, telomere length, and other stuff that I'm less interested in (OMICAge, extrincic epigenetic age-Hannum, as examples).
@@conqueragingordietrying1797 Thanks for the info. So, do they give you a discount for ordering so many tests per year? That certainly adds up. I guess I will start with one and we will see.
Okay, so food does not approach supplements for concentration of interesting science study ingredients and bioavailability.
For me, it's the opposite, but no worries, I appreciate other approaches, as long as biomarkers improve, and aging is slowed!
Depressing that protein is such a negative correlation with epigenetics. Seems counterintuitive that a strong individual who builds muscle is worse off in the long run, but that seems to be the case. I guess the ideal practice for longevity is as close to starvation as still maintains essential nutrients :(
Note that it could be a total fat, SFA, and coconut butter story, not protein, as these are correlations. It will take a lot more tests to sort out causation...
Current protein intake is ~100g/d, which is close to the 1.6g/kg BW that's been shown to maximize muscle mass in response to strength training-I'll make a small cut, based on the data, ~5-10g/d.
Methionine in animal protein is the precursor of the methyl groups which arenused in DNA methylation. There is already research which shows that methionine restriction elongates life. Look up recent 2024 Nurses prospective cohort study (korat et al) on protein intake and healthy aging. Plant protein was the stellar performer, animal protein not so much. Plant protein is lower on methionine.
The diet is rich in plant proteins, far less on animal (only sardines, for now)...
@@conqueragingordietrying1797 Yes, I have been following you. Do look up the Nurses cohort study though. The Horvath results do suggest excess DNA methylation, and methionine reduction should help. I do recall than sardines were your #1 calorie source.
Shouldnt you be using biomarkers based on healthy people? High sfa fat would be more problematic for someone who is obese and has heart disease, then for someone who is healthy, since they may want to increase their ldl.
The studies that I'm basing aging and all-cause mortality data are very large, hundreds of thousands of people (or more), which are a good reflection of general population averages.
If you know of longitudinal studies with "healthy" people's biomarkers changing during aging or mortality risk that disagrees with those studies, please post them...
Also, in terms of what may be problematic, rather than what might be true, I follow the data, which suggests lowering my protein, total fat, SFA, and coconut butter intake. Whether that makes a positive impact on biomarkers (or not), we'll see...
@@conqueragingordietrying1797 What looks healthy from giant data sets will be what helps unhealthy people more so then what helps healthy people. I am not an expert in this field at all, but I just wanted to point out a flaw in following whole population data for someone trying to maximize health to the extreme instead of simply avoid obesity, which is probably what would be the most important biomarker for unhealthy people, but completely irrelevant to you. So Horvarth may be way off.
If the non-Horvath clocks are wildly inaccurate and misleading for measuring the concept of "biological age" then why use them at all for that purpose. Are not people just fooling themselves and others when using the non-horvath clocks?
Hi @jackbuaer3828, not all-gold standard clocks include Horvath, but also DunedinPACE, PhenoAge, and GrimAge, as they're consistently associated with health-related outcomes in many studies.
Outside of those, you're right, buyer beware...
@@conqueragingordietrying1797If Pheonage is giving you 16 years younger and Horvarth is giving you 7 years older, then one or both of those clocks must be wildly inaccurate for the concept of measuring biological age.
If Horvarth is the more accurate biological age clock, that would suggest Phenoage is wildly inaccurate and misleading in terms of biological age.
Perhaps Phenoage is good at measuring risk reduction in terms of disease rather than measuring someone's relative biological age. I can see it as a useful tool for measuring risk reduction, but reducing one's risk of getting heart disease, cancer, and other conditions associated with aging is wildly different than someone having the biology of a person that is 16 years younger than their chronological age. I think this is why the use of biological clocks is confusing and misleading people.
If a biological clock is not good at measuring someone's biological age, perhaps it should give a health score rather than an age score.
I would not optimize horvath at all. I would optimize things like DunedInPace and GrimAge.
Hmm, I though about that, but it's the best clock for predicting chronological age. It's obviously not as good as the others you mentioned for ACM risk, but those tests aren't anywhere close to being as good as predicting chronological age.
@@conqueragingordietrying1797 Ideally, we'd optimised each one to get the most information from the murky epigenome algorithms. What the hierarchy of optimisation should be is certainly an interesting debate. BJ has done them all so it seems possible.