#1

Not good for TruD, but they probably use the chronological age as part of their calculations. That IMO is not ideal, but companies often do.

8 Likes
#2

Tell me Iā€™m wrong, but my general gestalt is that these age test are of little value and questionable accuracy. My logical side thinks how do you prove that these tests are accurate? Wait until someone dies and see if indeed they lived longer?

6 Likes
#3

well, whether or not they predict a real biological age, I think that they can deliver some valuable insights.
For example, took the grim age test and the greatest predictor of my mortality was vascular health. Since I mountain bike and feel great with BP of 104/68 I thought that this would have been my lowest chance of death. My Oura ring was also indicating this, but thought that was fake data, so now I have something to focus on.

2 Likes
#4

One methylation test showed me younger another from the same company showed me significantly older than my chronological age. One test showed my rate of aging considerably lower than 1 the next from the same company considerably above 1. The tests were not that far apart in time and I have stuck to a consistent health regimen.

The tests with poor outcome made me feel bad but now I donā€™t care there is a lot of noise as Matt Kaberlein says. I go for my biomarkers that with two exceptions (including genetically driven high lpa) are consistently good and in accordance with my regimen.

3 Likes
#5

He doesnā€™t mention what algorithm is used by each company? Thatā€™s important information.

On the other hand the fact the same sample and company tested twice gave such different results is very worrying.

I believe quality control is the biggest issue with epigenetic testingā€¦

2 Likes
#6

Hereā€™s an idea for a competitor:

Buy test in retail stores or everywhere pretty much, comes with a code. Take the test and mail it in. Use the code to check the results.

Thatā€™s actually more reliable then, no other data included like chronological age or lifestyle that can make it seem to the customer itā€™s B.S. Of course along with that have a great machine learning model or something to estimate the biological age.

#7

Not to defend any of these companies, but there may be legitimate reasons to ask for your calendar age. The exact same biomarker value may mean completely different things depending on whether youā€™re 5, 55, or 105 years of age, so they may need that info to evaluate.

That said, I donā€™t think thereā€™s any value in these tests, because the results returned are essentially random - completely all over the place and not even necessarily consistent internally from the same sample submitted separately.

Can these eventually with further work in the future provide some value or actionable advice, I donā€™t know, but as of today I personally see no point in it.

A bigger problem is that weā€™re all individuals N=1, and especially those of us focused on health and probably taking a variety of powerful drugs and supplements. We very likely are several standard deviations away from the general public in many respects. Meanwhile these tests go off of ranges based on the general population. I see this as very dubious for any single individual, and especially if youā€™re in an outlier subgroup.

For now, Iā€™ll pass, and furthermore I ignore all conclusions based on ā€œbecause my age test said xā€. There has been no verified validation to these tests, so meh. Entertaining at best. YMMV.

1 Like
#8

Fascinatingly, as part of my longevity medicine intake, Epigenetic Age is never discussed as something Iā€™d want to test. My wife and I did 3 different types, and we ranged from 8-9 years younger than chronological, to 16 years older. So whatā€™s the error of 25 years? Taking that and deciding on your n=1 maneuvers with such discrepant results ā€¦ good luck. Iā€™ll save my $$ and get a free NOVOS Faceage ā€¦ happy with those results - and they are consistent.

9 Likes
#9

ā€¦ or maybe an artifact to worry senselessly about.

2 Likes
#10

Yes - given results like those described by Matt, and the general lack of information from the companies producing these tests, I can see little value in them. Iā€™m happy to sit this out until some good validation data becomes available.

4 Likes
#11

GlycanAge is good for measuring your inflammation.

Consistent having used it for 3 years.

#12

I would presume that how the age prediction is made is a secret sauce and something he could disclose even if he knew. Also if prediction uses AI, there wouldnā€™t of course be an algorithm used but instead there is a trained model.

#13

Most of these epigenetic age clocks are licensed from academic groups (e.g. Steve Horvath, or Yale university) where they are developed. The companies themselves do not have the expertise to develop these clocks. Of course, how they implement them in software and the lab processes they use to actually measure them, are in the companyā€™s control but the companies provide no information on how they test and validate that their implementations are accurate or consistent. And this is completely separate from the broader issue of what exactly they are measuring and how it reflects on ā€œtrueā€ biological age (which is not something that has been validated yet, even in animals).

I still think this is something (the test validation) that needs to be done by the NIA ITP program as they have the largest database of animal longevity and ongoing studies, so would be a natural place to start.

#14

Great idea with that NIA ITP independent validation. Iā€™d love to see it happen. That said on practical side who would pay for such a validation? One of the possible approaches would be for the companies themselves to pay for it but it may cause conflict of interest.

#15

It would not be a direct test of the company epigenetic clocks as those are designed for humans, so theyā€™d have to be adjusted based on mouse epigenetic data, so this would have to be funded by the NIA or other non-profit group. It would be a startā€¦ if the clocks can be shown to accurately estimate the healthspan and/or lifespan of the model organisms then at least we have a starting point for the human clock validation. We are still a long way away from this right now.

1 Like
#16

Right. Also talking about various epigenetic tests as they appear to be measuring different things, I wonder if based on intervention, there may be a mapping to epigenetic clock that best captures the intervention results. You still use all the clocks during testing but you weigh the input from certain clocks as less important. In the end the set of successful interventions may systemically reverse the aging and move all epigenetic clocks backwards?

#17

I refuse to put money in the pockets of these age clock companies. I donā€™t believe any of them.

5 Likes
#18

Exactly how I feel but I was afraid to say it and piss off everyone. lol

1 Like
#19

Could be, but the test detected high levels of PAI-1 an vascular inflammation indicator which I was not aware of.