• Log in with Facebook Log in with Twitter Log In with Google      Sign In    
  • Create Account
  LongeCity
              Advocacy & Research for Unlimited Lifespans

Photo

Biological Age

anti-aging

  • Please log in to reply
143 replies to this topic
⌛⇒ write a quiz!

#121 Iporuru

  • Guest
  • 119 posts
  • 124
  • Location:Europe

Posted 23 January 2019 - 05:43 PM

DNA methylation GrimAge strongly predicts lifespan and healthspan.

Lu AT, Quach A, Wilson JG, Reiner AP, Aviv A, Raj K, Hou L, Baccarelli AA, Li Y, Stewart JD, Whitsel EA, Assimes TL, Ferrucci L, Horvath S.
Aging (Albany NY). 2019 Jan 21. doi: 10.18632/aging.101684. [Epub ahead of print]
PMID: 30669119

https://s3-us-west-1...cHAzptsYSEH.pdf

 

Abstract
It was unknown whether plasma protein levels can be estimated based on DNA methylation (DNAm) levels, and if so, how the resulting surrogates can be consolidated into a powerful predictor of lifespan. We present here, seven DNAm-based estimators of plasma proteins including those of plasminogen activator inhibitor 1 (PAI-1) and growth differentiation factor 15. The resulting predictor of lifespan, DNAm GrimAge (in units of years), is a composite biomarker based on the seven DNAm surrogates and a DNAm-based estimator of smoking pack-years. Adjusting DNAm GrimAge for chronological age generated novel measure of epigenetic age acceleration, AgeAccelGrim.Using large scale validation data from thousands of individuals, we demonstrate that DNAm GrimAge stands out among existing epigenetic clocks in terms of its predictive ability for time-to-death (Cox regression P=2.0E-75), time-to-coronary heart disease (Cox P=6.2E-24), time-to-cancer (P= 1.3E-12), its strong relationship with computed tomography data for fatty liver/excess visceral fat, and age-at-menopause (P=1.6E-12). AgeAccelGrim is strongly associated with a host of age-related conditions including comorbidity count (P=3.45E-17). Similarly, age-adjusted DNAm PAI-1 levels are associated with lifespan (P=5.4E-28), comorbidity count (P= 7.3E-56) and type 2 diabetes (P=2.0E-26). These DNAm-based biomarkers show the expected relationship with lifestyle factors including healthy diet and educational attainment.Overall, these epigenetic biomarkers are expected to find many applications including human anti-aging studies.
 


  • Informative x 4

#122 QuestforLife

  • Member
  • 541 posts
  • 230
  • Location:UK

Posted 24 January 2019 - 02:45 PM

Pretty mindblowing paper!


  • Agree x 2

Click HERE to rent this BIOSCIENCE adspot to support LongeCity (this will replace the google ad above).

#123 HighDesertWizard

  • Member
  • 734 posts
  • 439
  • Location:Bend, Oregon, USA

Posted 25 January 2019 - 09:10 PM

Pretty mindblowing paper!

 

I agree.

  • This study describes an approach to prediction of Survival Probability Odds that is capable of predicting the greatest difference between high and low odds of future survival based on Protein Surrogates found in the blood.

 

Supplementary Figure 15. Kaplan Meier plots of individuals who age slowly/quickly according to different measures of epigenetic age acceleration.

 

Each panel depicts two Kaplan Meier plots for survival distributions, i.e. the probability of being alive (y-axis) at a given time after the blood draw (x-axis). The two plots/lines correspond to two groups of individuals: fast epigenetic agers (defined as being in the top 20% percentile of epigenetic age acceleration) and slow epigenetic agers (defined to be in the bottom 20% percentile). Each row reports the results for a different measure of epigenetic age acceleration: first row (A-E) AgeAccelGrim, second row (F-J) AgeAccelPheno, third row (K-O) AgeAccelHannum (based on Hannum 2013), fourth row (P-T) AgeAccelerationResidual (based on Horvath 2013). Columns correspond to the different data sets (corresponding to sub-studies within cohorts). The results cross studies were combined by Stouffer’s method. When it comes to detecting a different mortality risk between fast and slow agers, AgeAccelGrim has the best performance (Stouffer's meta analysis P=6.4E-38), followed by AgeAccelPheno (P=5.7E-21), AgeAccelHannum (P= 1.3E-5) and AgeAccelResidual (P=0.17).

 

V3gZ8Lsh.png



sponsored ad

  • Advert
Advertisements help to support the work of this non-profit organisation. [] To go ad-free join as a Member.

#124 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 26 January 2019 - 02:25 PM

Pretty mindblowing paper!

 

Yes, it is definitively a very impressive piece of work and it would require a huge amount of time to fully understand the details, at least for me.

 

I also wonder how actionable it is for us. I would greatly appreciate your comments on this !

 

I found a bit more digestible and actionable the DNAm PhenoAge and Phenotypic Age calculator by Levine and few of the same authors as also discussed in this thread, e.g. here and here. The plasma biomarkers in there are more commonly used in the clinic and I could quite easily calculate my Phenotypic Age.

 

The less commonly measured but very important biomarkers such as GDF-15 (see also the chart I posted in this thread) and PAI‐1 stand out. That is important but I do not have them allowing me to track the AgeAccelGrim.

 

Moreover, the authors rightly say: "...Our DNAm-based surrogate biomarkers of plasma protein levels may be leveraged by researchers who rely on bio-banked DNA samples without the availability of plasma samples. Strong evidence supports links between plasma proteins used in the construction of GrimAge and various age-related conditions ....” That is great but I would also be interested to the reverse problem as I lack my DNA methylation data, while maintaining a pretty good log of more common clinic biomarkers.

 

I will continue to study the paper but after lot of reading I start to be quite impatient to see more work focusing on one single approach to biomarkers of aging and its calibration against intervention where we do know (e.g. epigenetics rejuvenation?) aging hallmarks are regressed so I can compare with my data. I am pretty sure it will come, possibly from the same authors, maybe the self-tracking community or some other top researchers (I have confirmation from one of them) owing vast amount of personal omics and clinical longitudinal data.


  • like x 1

#125 QuestforLife

  • Member
  • 541 posts
  • 230
  • Location:UK

Posted 27 January 2019 - 08:16 AM

Yes, it is definitively a very impressive piece of work and it would require a huge amount of time to fully understand the details, at least for me.

I also wonder how actionable it is for us. I would greatly appreciate your comments on this !

I found a bit more digestible and actionable the DNAm PhenoAge and Phenotypic Age calculator by Levine and few of the same authors as also discussed in this thread, e.g. here and here. The plasma biomarkers in there are more commonly used in the clinic and I could quite easily calculate my Phenotypic Age.

The less commonly measured but very important biomarkers such as GDF-15 (see also the chart I posted in this thread) and PAI‐1 stand out. That is important but I do not have them allowing me to track the AgeAccelGrim.

Moreover, the authors rightly say: "...Our DNAm-based surrogate biomarkers of plasma protein levels may be leveraged by researchers who rely on bio-banked DNA samples without the availability of plasma samples. Strong evidence supports links between plasma proteins used in the construction of GrimAge and various age-related conditions ....” That is great but I would also be interested to the reverse problem as I lack my DNA methylation data, while maintaining a pretty good log of more common clinic biomarkers.

I will continue to study the paper but after lot of reading I start to be quite impatient to see more work focusing on one single approach to biomarkers of aging and its calibration against intervention where we do know (e.g. epigenetics rejuvenation?) aging hallmarks are regressed so I can compare with my data. I am pretty sure it will come, possibly from the same authors, maybe the self-tracking community or some other top researchers (I have confirmation from one of them) owing vast amount of personal omics and clinical longitudinal data.


Yes, it's complicated, but I guess the takeaway is that we can get a good idea of how we are aging from a selection of (mostly) common blood tests.

It's still not clear how methylation changes are related to the blood changes, whether they are up or downstream.

From my point of view the jury is still out on the relationship with telomeres, which was once assumed to be a correlated relationship from the original Horvath's clock (i.e. longer telomeres keeping a cell line for longer meant older epigenetic age), but now appears in the later clocks to be something more complex or inverse with shorter telomeres also leading to an older epigenetic age (perhaps relating to the inability of short telomeres to turn over cells).

I'm sure we'll see lots more papers like this soon, as it's not difficult to do statistics on blood markers and methylation from large blood banks.
  • like x 1

#126 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 27 January 2019 - 10:37 AM

Does someone here know about a forum, blog or other platforms, also professional, specifically dedicated to the investigation of biomarkers of aging? I would be interested to follow a more dedicated platform, possibly reporting also interesting finding in this thread. The information is scattered all around.



Click HERE to rent this BIOSCIENCE adspot to support LongeCity (this will replace the google ad above).

#127 QuestforLife

  • Member
  • 541 posts
  • 230
  • Location:UK

Posted 27 January 2019 - 10:44 AM

Does someone here know about a forum, blog or other platforms, also professional, specifically dedicated to the investigation of biomarkers of aging? I would be interested to follow a more dedicated platform, possibly reporting also interesting finding in this thread. The information is scattered all around.


You mean apart from the one on this site? Longecity ran a biomarkers thread for members last year where they contributed to the cost of telomere and epigenetic agig tests. It's still in its infancy but will hopefully continue and expand in future years.

#128 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 27 January 2019 - 12:15 PM

You mean apart from the one on this site? Longecity ran a biomarkers thread for members last year where they contributed to the cost of telomere and epigenetic agig tests. It's still in its infancy but will hopefully continue and expand in future years.

 

Yes thank you. I knew about it but did not participated. I also do hope this continues (there is also an effort led by Josh Mitteldorf if you follow his blog) and I am eager to see results from this particular community. I meant about a place where the current research is discussed and scrutinized by professionals, industrialists and scientists more deeply involved. Or maybe a recurring conference dedicated to the topic? 
 



#129 QuestforLife

  • Member
  • 541 posts
  • 230
  • Location:UK

Posted 27 January 2019 - 06:30 PM

Yes thank you. I knew about it but did not participated. I also do hope this continues (there is also an effort led by Josh Mitteldorf if you follow his blog) and I am eager to see results from this particular community. I meant about a place where the current research is discussed and scrutinized by professionals, industrialists and scientists more deeply involved. Or maybe a recurring conference dedicated to the topic?


Read it?

https://www.liebertp...9/rej.2018.2083

(Need sci-hub)

Seems to me the science is moving so fast you'd want to use the very latest methylation clock, but I guess it doesn't matter once you've collected the blood sample. The weakness of Josh's idea is you need a lot of participants.

#130 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 27 January 2019 - 08:38 PM

Read it?

https://www.liebertp...9/rej.2018.2083

(Need sci-hub)

Seems to me the science is moving so fast you'd want to use the very latest methylation clock, but I guess it doesn't matter once you've collected the blood sample. The weakness of Josh's idea is you need a lot of participants.

 

Absolutely, yes, I have seen it. Thank you :-)
 



#131 HighDesertWizard

  • Member
  • 734 posts
  • 439
  • Location:Bend, Oregon, USA

Posted 28 January 2019 - 12:10 AM

Yes, it is definitively a very impressive piece of work and it would require a huge amount of time to fully understand the details, at least for me.
 
I also wonder how actionable it is for us. I would greatly appreciate your comments on this !
 
I found a bit more digestible and actionable the DNAm PhenoAge and Phenotypic Age calculator by Levine and few of the same authors as also discussed in this thread, e.g. here and here. The plasma biomarkers in there are more commonly used in the clinic and I could quite easily calculate my Phenotypic Age.
 
The less commonly measured but very important biomarkers such as GDF-15 (see also the chart I posted in this thread) and PAI‐1 stand out. That is important but I do not have them allowing me to track the AgeAccelGrim.
 
Moreover, the authors rightly say: "...Our DNAm-based surrogate biomarkers of plasma protein levels may be leveraged by researchers who rely on bio-banked DNA samples without the availability of plasma samples. Strong evidence supports links between plasma proteins used in the construction of GrimAge and various age-related conditions ....” That is great but I would also be interested to the reverse problem as I lack my DNA methylation data, while maintaining a pretty good log of more common clinic biomarkers.
 
I will continue to study the paper but after lot of reading I start to be quite impatient to see more work focusing on one single approach to biomarkers of aging and its calibration against intervention where we do know (e.g. epigenetics rejuvenation?) aging hallmarks are regressed so I can compare with my data. I am pretty sure it will come, possibly from the same authors, maybe the self-tracking community or some other top researchers (I have confirmation from one of them) owing vast amount of personal omics and clinical longitudinal data.

 
I appreciate your posts albedo...
 
I have a few alternative perspectives about the value and implications of the GrimAge study. I'll focus here on more general comments and, in a follow-up post, on the, truly, profound and actionable, insight the GrimAge Study provides.

  • I disagree that it would be a good thing "to see more work focusing on one single approach" to biomarkers of aging.
     
  • As stated other places, I'm a Fallibilist. I believe it's critical that we be skeptical about any fixed knowledge that 'we all know" is true. This aging-biomarkers field is in its infancy. It's in our interest that the techniques evolve.
     
  • The latest studies coming out of Horvath's lab are focusing on the Machine Learning Knowledge Target Remaining Lifespan rather than Biological Age. In my view, this is a great change.

This is a development I anticipated 18 months ago in a thread called Longevity Velocity and how to Calculate it.

 

But in that thread, I had imagined that the findings of studies focused on Remaining Lifespan would provide an algorithm returning a single number of years / months remaining. I see now that studies focused on Remaining Lifespan rather than Biological Age will be providing a set of Survival Probability Curves for various demographic- and health-related categories.

 

This, in itself, IMO, is a good thing. And it's good also that the GrimAge Survival Probability findings distinguish Survival Probability curves for two groups with greater variance than ever before.

 

That is a cause for celebration because, as I'll show in my next post, a lot is known about key independent variables driving PAI-1 and GDF-15 increase. And, well, it seems Mother Nature, herself, has been focused on those variables.

 

:) 
 

 

V3gZ8Lsh.png

 

 

  • It's a good thing, also, because Life Insurance industry Actuary pioneers are focused on Remaining Lifespan as a useful independent variable for predicting Remaining Life risk.

The good news... The Life Insurance Industry will be funding these kinds of studies.

 

YouSurance is First to Assess Health and Lifespan Using Epigenetic Biomarkers for Life Insurance Applicants

 

Life Epigenetics


Edited by HighDesertWizard, 28 January 2019 - 12:20 AM.


#132 HighDesertWizard

  • Member
  • 734 posts
  • 439
  • Location:Bend, Oregon, USA

Posted 28 January 2019 - 10:52 AM

I also wonder how actionable it is for us. I would greatly appreciate your comments on this !

 
I've found this study important also just because the content of its biological independent variable findings are actionable.
 
I'm a Fallibiist. And that means significant criticism of my key beliefs, both direct and implicit, ought to command my attention. And it does because taking on criticism directly is required for making Fast Progress. Facing criticism.

  • I've been posting about The Inflammatory Reflex for almost 7 years now. I've been posting about Survival Probability Curves (aka, Kaplan-Meier Curves) for 4 years now. And the Survival Probability Curves associated with the means to trigger it are the most significant I've ever seen. (If you've seen better, please post so we can discuss.)
     
  • The Inflammatory Reflex is important also because it is a biological mechanism with unique anatomy within us that would have required profound and systematic attention from Mother Nature over a million centuries*.
    • I say "a million centuries" because we share key biological functions of TIR with rodents and, while there is debate about how long ago we share an ancestor with rodents, 70 million appears to be the shortest estimated period. And I figure that it would have taken a few 10s of millions for the essential vagus nerve functions we share with rodents related to The Inflammatory Reflex to develop. Hence, I believe "a million centuries" is a roughly accurate length of time that evolution has been establishing the mechanism now called The Inflammatory Reflex.
  • All that said, this GrimAge study comes along with Survival Curves demonstrating that a few key proteins and a Machine Learning based Algorithm can predict variances between high- and low-probability survivors. IMO, that's a big deal.

And because I'm a Fallibilist, I must chase down the facts about the independent variables implicated to determine whether they are related to the mechanism of The Inflammatory Reflex. If these GrimAGE variables aren't related to Inflammatory Reflex variables, I'll need to rethink how important I've believed The Inflammatory Reflex is.

 

That's the background for my digging into the details a bit.

 
Two independent variables popped up as important for the investigators, Plasminogen Activator Inhibitor 1 (PAI-1) and Growth differentiation factor-15 (GDF-15.

Is there an intersection of the variables driving the increase of PAI-1 and GDF15 and the effects of The Inflammatory Reflex?

If there is not, I need to rethink how important The Inflamamtory Reflex is.

albedo... If there is, perhaps you might consider becoming more familiar with The Inflammatory Reflex yourself and how to trigger it.

 
Question: What is the key measurable. biological variable implicated by The Inflammatory Reflex as correlated with disease and death?
 
Answer: Higher TNF Expression in the Spleen and in the Serum.

 

 
Maximum survival probability benefit of triggering The Inflammatory Reflex requires 5 things...

  • Intact Vagus Nerve Signaling
  • Intact Splenic Nerve Signaling
  • Intact Spleen
  • Functional and not Antagonized Muscarinic Receptors in Brain
  • a7 Nicotinic-Acetylcholine Receptor Functionality
These 5 physiological/biological requirements for a viable Inflammatory Reflex are highlighted in the pics below highlighted in green.
 
gallery_16949_60_173943.jpg
 
 

 

gallery_16949_60_196876.jpg
 
 
Notice that Serum TNF is reduced proportionately greater than Splenic TNF in the Electro-Acupuncture Experiment. A comparable experiment comparing the two was also performed in the Xanomeline study. Here's the pic...

9jm0hC6h.png

 

 

DOlhIjK.png
 
 

 
 
Does TNF expression increase PAI-1?
 
PAI-1

Plasminogen activator inhibitor type 1 (PAI-1) is induced by many proinflammatory and pro-oxidant factors. Among them, tumor necrosis factor alpha (TNFalpha), a pivotal early mediator that regulates and amplifies the development of inflammation, is one of the strongest PAI-1 synthesis activators.

 

  
The finding that PAI-1 is important in the GrimAGE study is a kind of confirmation that Mother Nature knew what she was doing in establishing The Inflammatory Reflex.

 

albedo... In fact, the GrimAGE study emphasized a single most important biological variable, PAI-1, that is reduced by the most significant in-the-serum-molecular-target of The Inflammatory Reflex. This knowledge is Actionable.

 

From time to time, my practice of triggering The Inflammatory Reflex gets sloppy. But the content findings of this GrimAGE study are motivating and I'll be focusing on increasing discipline in implementing my regimen focused on triggering The Inflammatory Reflex.

 

:)


Edited by HighDesertWizard, 28 January 2019 - 11:30 AM.

  • like x 1

#133 HighDesertWizard

  • Member
  • 734 posts
  • 439
  • Location:Bend, Oregon, USA

Posted 28 January 2019 - 12:57 PM

 Does TNF expression increase PAI-1?

 
PAI-1

Plasminogen activator inhibitor type 1 (PAI-1) is induced by many proinflammatory and pro-oxidant factors. Among them, tumor necrosis factor alpha (TNFalpha), a pivotal early mediator that regulates and amplifies the development of inflammation, is one of the strongest PAI-1 synthesis activators.

 

I misplaced a reference to a significant study about the relationship of TNF to PAI-1 and have now found it.

 

TNF-alpha, but not IL-6, stimulates plasminogen activator inhibitor-1 expression in human subcutaneous adipose tissue

 

Reducing PAI-1... Actionable, via The Inflammatory Reflex.

 

:)


  • Informative x 2

#134 QuestforLife

  • Member
  • 541 posts
  • 230
  • Location:UK

Posted 28 January 2019 - 01:22 PM

It is likely that PAI-1 up regulation is related to LDL in the blood, i.e. decreasing clotting in order to reduce the chance of a blockage (cardiovascular disease). There is some support for this view:

 

https://www.ncbi.nlm...pubmed/28692480

 

 

 

The plasma PAI-1 levels may be determined by the degree of obesity and TG metabolic disorders. These factors were also shown to be correlated with a decreased LDL-particle size, increasing the risk of ASCVD, even in nondiabetic patients with well-controlled serum LDL-C levels

 

The connection with inflammatory reflex might not be immediately apparent, but it is known LDL is anti-infection (possibly entering via gut) and as far as I can tell, the inflammatory reflex is a reaction of the body to an infection, or something that looks like one.

 

 


Edited by QuestforLife, 28 January 2019 - 01:24 PM.


#135 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 29 January 2019 - 10:22 AM

In case you overlooked this, it is 7.5 k$ though! I expect some clinics will buy it and re-sell as a service. I recollect in a couple of occasion having used the H-Scan predecessor.

https://www.agemeter..._eid=44a6ef5d7b

 

Witnessing and re-emphasizing the importance of functional biomarkers, also this article on the AgeMeter is interesting. Notice the endorsements on this type of measurement, next to molecular biomarkers and omics, by Aubrey deGrey, George Church and David Sinclair:

AgeMeter: Physiological Biomarkers to Determine Functional Age

 



#136 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 29 January 2019 - 02:43 PM

 
I appreciate your posts albedo...
 
I have a few alternative perspectives about the value and implications of the GrimAge study. I'll focus here on more general comments and, in a follow-up post, on the, truly, profound and actionable, insight the GrimAge Study provides.

  • I disagree that it would be a good thing "to see more work focusing on one single approach" to biomarkers of aging.
     
  • As stated other places, I'm a Fallibilist. I believe it's critical that we be skeptical about any fixed knowledge that 'we all know" is true. This aging-biomarkers field is in its infancy. It's in our interest that the techniques evolve.
     
  • The latest studies coming out of Horvath's lab are focusing on the Machine Learning Knowledge Target Remaining Lifespan rather than Biological Age. In my view, this is a great change.

This is a development I anticipated 18 months ago in a thread called Longevity Velocity and how to Calculate it.

 

But in that thread, I had imagined that the findings of studies focused on Remaining Lifespan would provide an algorithm returning a single number of years / months remaining. I see now that studies focused on Remaining Lifespan rather than Biological Age will be providing a set of Survival Probability Curves for various demographic- and health-related categories.

 

This, in itself, IMO, is a good thing. And it's good also that the GrimAge Survival Probability findings distinguish Survival Probability curves for two groups with greater variance than ever before.

 

That is a cause for celebration because, as I'll show in my next post, a lot is known about key independent variables driving PAI-1 and GDF-15 increase. And, well, it seems Mother Nature, herself, has been focused on those variables.

 

:) 
 

 

  • It's a good thing, also, because Life Insurance industry Actuary pioneers are focused on Remaining Lifespan as a useful independent variable for predicting Remaining Life risk.

The good news... The Life Insurance Industry will be funding these kinds of studies.

 

YouSurance is First to Assess Health and Lifespan Using Epigenetic Biomarkers for Life Insurance Applicants

 

Life Epigenetics

 

I am impressed and must admit a bit overwhelmed by this and following post knowledge. It confirms this Forum is a good place to keep learning :)

 

The GrimAge paper is fantastic but I maintain my impatience re the "single approach". Far to stop research of better and better clocks I only mean that, as we are reaching a point now to have means, at least in principle, to regress most if not all the hallmarks of aging (say à-la-OSKM or small molecules or EV's, please refer to the OSKM thread you are greatly contributing to) I wonder what it would take to focus on one of the available clocks and assess how its rate slows down with the intervention. This is a calibration phase which hopefully can be then translated to a similar clock applied to humans using the same multivariate linear approach. Or maybe directly using CR or exercise or small molecules or other interventions on individuals monitored over time which, when compared to animal models, is a longer, more expensive and more complex process. But again I am pretty sure this will come ....

 

I hope I understand your point re lifespan and biological age and your focus on lifespan and the K-M curves. That is the standard way of assessment. But also the Levine's clock does that. I only felt its construction is based, in Step 1, on a regression which selected a number of biomarkers (from the 42 available in the analyzed cohorts to 9) much more in use in the clinical practice than those used in GrimAge such as Plasminogen Activator Inhibitor 1 (PAI-1) and the Growth differentiation factor-15 (GDF-15). So I felt that clock is more useful to me right now. However, I do agree PAI-1 and GDF-15 can point even more precisely to underlying biochemistry of aging and theories such as the Inflammatory Reflex you are emphasizing. I must admit I am more focused these days to biological age to see if what I am trying on myself is detectable and can motivate further research than following theories of aging.

 

 

 
I've found this study important also just because the content of its biological independent variable findings are actionable.
 
I'm a Fallibiist. And that means significant criticism of my key beliefs, both direct and implicit, ought to command my attention. And it does because taking on criticism directly is required for making Fast Progress. Facing criticism.

  • I've been posting about The Inflammatory Reflex for almost 7 years now. I've been posting about Survival Probability Curves (aka, Kaplan-Meier Curves) for 4 years now. And the Survival Probability Curves associated with the means to trigger it are the most significant I've ever seen. (If you've seen better, please post so we can discuss.)
     
  • The Inflammatory Reflex is important also because it is a biological mechanism with unique anatomy within us that would have required profound and systematic attention from Mother Nature over a million centuries*.
    • I say "a million centuries" because we share key biological functions of TIR with rodents and, while there is debate about how long ago we share an ancestor with rodents, 70 million appears to be the shortest estimated period. And I figure that it would have taken a few 10s of millions for the essential vagus nerve functions we share with rodents related to The Inflammatory Reflex to develop. Hence, I believe "a million centuries" is a roughly accurate length of time that evolution has been establishing the mechanism now called The Inflammatory Reflex.
  • All that said, this GrimAge study comes along with Survival Curves demonstrating that a few key proteins and a Machine Learning based Algorithm can predict variances between high- and low-probability survivors. IMO, that's a big deal.

And because I'm a Fallibilist, I must chase down the facts about the independent variables implicated to determine whether they are related to the mechanism of The Inflammatory Reflex. If these GrimAGE variables aren't related to Inflammatory Reflex variables, I'll need to rethink how important I've believed The Inflammatory Reflex is.

 

That's the background for my digging into the details a bit.

 
Two independent variables popped up as important for the investigators, Plasminogen Activator Inhibitor 1 (PAI-1) and Growth differentiation factor-15 (GDF-15.

Is there an intersection of the variables driving the increase of PAI-1 and GDF15 and the effects of The Inflammatory Reflex?

If there is not, I need to rethink how important The Inflamamtory Reflex is.

albedo... If there is, perhaps you might consider becoming more familiar with The Inflammatory Reflex yourself and how to trigger it.

 
Question: What is the key measurable. biological variable implicated by The Inflammatory Reflex as correlated with disease and death?
 
Answer: Higher TNF Expression in the Spleen and in the Serum.

 

 

 
 
Does TNF expression increase PAI-1?
 
PAI-1

Plasminogen activator inhibitor type 1 (PAI-1) is induced by many proinflammatory and pro-oxidant factors. Among them, tumor necrosis factor alpha (TNFalpha), a pivotal early mediator that regulates and amplifies the development of inflammation, is one of the strongest PAI-1 synthesis activators.

 

  
The finding that PAI-1 is important in the GrimAGE study is a kind of confirmation that Mother Nature knew what she was doing in establishing The Inflammatory Reflex.

 

albedo... In fact, the GrimAGE study emphasized a single most important biological variable, PAI-1, that is reduced by the most significant in-the-serum-molecular-target of The Inflammatory Reflex. This knowledge is Actionable.

 

From time to time, my practice of triggering The Inflammatory Reflex gets sloppy. But the content findings of this GrimAGE study are motivating and I'll be focusing on increasing discipline in implementing my regimen focused on triggering The Inflammatory Reflex.

 

:)

 

This is a very interesting part of your post. Coincidentally I have requested an inflammation panel for a condition I have and included both TNF-alpha and IL-6. After several years of collaboration with my doctor he was willing to accept this addition to the more common chronic inflammation biomarkers even if it is hard to explain to insurances the clinical relevance vs. research. It is also a quite tricky field as, if the common internet wisdom is to lower TNF-alpha, very strong drug inhibitors (e.g. for autoimmune diseases) are absolutely not always recommended as increasing other risk factor (e.g. mycobacterium infections). I am supplementing with a lower dose of curcumin which might impact TNF-alpha but hopefully on the safe side. And then you have the links you point to TNF-alpha vs IL-6 impacting PAI-1. So yes I agree with you also GrimAge is also actionable :)

 

 

It is likely that PAI-1 up regulation is related to LDL in the blood, i.e. decreasing clotting in order to reduce the chance of a blockage (cardiovascular disease). There is some support for this view:

 

https://www.ncbi.nlm...pubmed/28692480

 

 

 

 

The connection with inflammatory reflex might not be immediately apparent, but it is known LDL is anti-infection (possibly entering via gut) and as far as I can tell, the inflammatory reflex is a reaction of the body to an infection, or something that looks like one.

 

Thank you. It is very likely not directly related but indirectly your comment is suggesting of LP(a) as a "proxy" of IL-6. Just a thought .... I posted about this here.



#137 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 01 February 2019 - 06:25 PM

This is quite interesting, waiting for publication:

 

Wood T, Kelly C, Roberts M and Walsh B. An interpretable machine learning model of biological age [version 1; referees: 1 approved with reservations]. F1000Research 2019, 8:17

https://doi.org/10.1...esearch.17555.1

 

Note the critic review by Dr Alex Zhavoronkov of Insilico Medicine which already developed aging.ai and young.ai using a AI/ML approach (DNN). We discussed aging.ai previously in this thread.

 

Anyone here who tried using the following commercial tool developed by the authors (disclaimed in the competing interests)? How would it compare with aging.ai?

https://home.bloodcalculator.com/

 

Finally I wonder to which extent the SHAP (Shapley additive explanations plots It looks to me this technique might ease the determination of the relative weights importance of the Levine's Phenotypic Age as discussed here. Indeed it looks to me the technique does just that: "...SHAP plots of input markers. SHAP summary plots (Figure 2) were used to determine which markers have the greatest influence on predicted biological age..."

https://doi.org/10.1...esearch.17555.1



#138 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 01 February 2019 - 08:28 PM

Sorry I pressed my last "post" button too fast. The last paragraph should read like this:

 

"Finally I wonder to which extent the SHAP (SHapley Additive exPlanations) plots might ease the determination of the relative weights importance of the Levine's Phenotypic Age as discussed here. Indeed it looks to me the technique does just that: "...SHAP plots of input markers. SHAP summary plots (Figure 2) were used to determine which markers have the greatest influence on predicted biological age...""

https://doi.org/10.1...esearch.17555.1



#139 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 11 February 2019 - 03:12 PM

Sorry I pressed my last "post" button too fast. The last paragraph should read like this:

 

"Finally I wonder to which extent the SHAP (SHapley Additive exPlanations) plots might ease the determination of the relative weights importance of the Levine's Phenotypic Age as discussed here. Indeed it looks to me the technique does just that: "...SHAP plots of input markers. SHAP summary plots (Figure 2) were used to determine which markers have the greatest influence on predicted biological age...""

https://doi.org/10.1...esearch.17555.1

 

Attached File  SHAP.PNG   249.26KB   0 downloads

 

Looking at the top 5, quite similar results in terms of relative importance of biomarkers to predicted age to those already published in 2016 by the team at Insilico Medicine:

Putin E, Mamoshina P, Aliper A, et al. Deep biomarkers of human aging: Application of deep neural networks to biomarker development. Aging (Albany NY). 2016;8(5):1021-33.



#140 aribadabar

  • Guest
  • 576 posts
  • 209
  • Location:Canada

Posted 12 February 2019 - 05:09 PM

attachicon.gif SHAP.PNG

 

Looking at the top 5, quite similar results in terms of relative importance of biomarkers to predicted age to those already published in 2016 by the team at Insilico Medicine:

Putin E, Mamoshina P, Aliper A, et al. Deep biomarkers of human aging: Application of deep neural networks to biomarker development. Aging (Albany NY). 2016;8(5):1021-33.

 

Interesting graphics.

 

Would you explain what the "thickness" and the "length" of the blot means, especially when it goes from high negative all the way to high positive value (e.g. BUN)?

I surmise, and correct me if I am wrong, blue/red denotes positive/negative impact on the model age calculation?

 

 

Looking at Fig. B, I have several questions:

- Looking at Albumin, what thin red negative vs, shorter but fatter bluish positive tails mean?

- What Trigs' long thin blue negative line denotes? Conversely, Lymphs (Absolute)  has long thin blue positive line?

- What about the long red negative in Phosphorous and the odd double-red Creatinine negative/positive line?

 

Thanks for deciphering!



#141 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 13 February 2019 - 05:37 PM

Interesting graphics.

 

Would you explain what the "thickness" and the "length" of the blot means, especially when it goes from high negative all the way to high positive value (e.g. BUN)?

...

 

I only wished to make the point that apparently different methods, e.g. when comparing to the Insilico's work with aging.ai, produced similar results (e.g. see albumin and glucose) than the subject recent but unpublished work which probably lacks real novelty.

 

It was the first time I have crossed SHAP plots and have no time to investigate further so I only rely on what they write.

 

So I understand it as following:

 

If I take BUN as an example, “red” means “increasing impact” on age predicted (e.g. on the x axis for men (B) say additional 10 years when comparing to chronological age) so it is a negative thing. OTC “blue” means “decreasing impact” on age predicted which is a positive thing. That should answer your “length” question. Wrt to the “thickness” question, that is I guess a variability in the distribution of the data set at a specific x. Consider also that the plots contains hundreds and thousands of dots which overlap.

 

...

I surmise, and correct me if I am wrong, blue/red denotes positive/negative impact on the model age calculation?

...

 

Yes, I think so.

 

...

- Looking at Albumin, what thin red negative vs, shorter but fatter bluish positive tails mean?

...

 

In the model, an increasing albumin has an extraordinary impact in the positive sense, meaning to reduce the predicted age say by 10 years, but that the data set variability is less than in the opposite side. A decreasing albumin has less impact in the negative sense, adding say up to 5-6 years to the predicted age, but the data set features much larger variability.

 

...

- What Trigs' long thin blue negative line denotes? Conversely, Lymphs (Absolute)  has long thin blue positive line?

...

 

In the model, decreasing Trigs decreases predicted age up to say 6-7 years with low variability. The negative effect on the predicted age is not similar to the positive effect and you also have higher variability.

 

Wrt lymphocytes (abs) it is quite known that a reduction is one of the most prominent signs of an immune system aging and I think the plot reflects that by adding years to the predicted age with low variability.

 

...

- What about the long red negative in Phosphorous and the odd double-red Creatinine negative/positive line?

...

 

See previous reply, mutatis mutandis.  

 

Wrt to creatinine, yes it looks a bit confusing. It might simply means that the model is not selective enough to disentangle the creatinine effect. Other models (e.g. Insilico’s) feature also a lower predictive impact of creatinine when compared to other biomarkers which of course does not mean that creatinine is not clinically important (kidney health). It might only means the model is not capturing the fullness of the age impact of the kidney function, as measured by creatinine, and would need to be improved. In the model BUN does a better job regarding the kidney health impact on age prediction.

 

In general, these models are age predictors and you need also to look at biological age predictors e.g. Phenotypic age or DNAm PhenoAge and even then I have questions as I pointed out in this thread (and also to the respective  authors) on the relative roles of biomarkers (e.g. CRP, RDW)

 

That is my interpretation. It might be worth to contact authors though. Let us know if you discover more :)

 

I hope this helps!


Edited by albedo, 13 February 2019 - 05:42 PM.

  • Informative x 2

#142 VP.

  • Member
  • 424 posts
  • 143

Posted 14 February 2019 - 07:28 PM

Uncovering a 'smoking gun' of biological aging clocks

A newly discovered ribosomal DNA (rDNA) clock can be used to accurately determine an individual's chronological and biological age, according to research led by Harvard T.H. Chan School of Public Health. The ribosomal clock is a novel biomarker of aging based on the rDNA, a segment of the genome that has previously been mechanistically linked to aging. The ribosomal clock has potentially wide applications, including measuring how exposures to certain pollutants or dietary interventions accelerate or slow aging in a diversity of species, including mice and humans.

 

"We have hopes that the ribosomal clock will provide new insights into the impact of the environment and personal choices on long-term health," said senior author Bernardo Lemos, associate professor of environmental epigenetics. "Determining biological age is a central step to understanding fundamental aspects of aging as well as developing tools to inform personal and public health choices."

The study will be published online in Genome Research on February 14, 2019.

Aging is exhibited by organisms as diverse as yeast, worms, flies, mice, and humans. Age is also the major risk factor for a plethora of diseases, including neurological diseases, cardiovascular diseases, and cancer. There are two types of age: chronological age, or the number of years a person or animal has lived, and biological age, which accounts for various lifestyle factors that can shorten or extend lifespan, including diet, exercise, and environmental exposures. Overall, biological age has been shown to be a better predictor of all-cause mortality and disease onset than chronological age.

For this new study, the researchers looked at the rDNA, the most active segment of the genome and one which has also been mechanistically linked to aging in a number of previous studies. Lemos and lead author Meng Wang, a research fellow in the Department of Environmental Health, hypothesized that the rDNA is a "smoking gun" in the genomic control of aging and might harbor a previously unrecognized clock. To explore this concept, they examined epigenetic chemical alterations (also known as DNA methylation) in CpG sites, where a cytosine nucleotide is followed by a guanine nucleotide. The study homed in on the rDNA, a small (13 kilobases) but essential and highly active segment of the genome, as a novel marker of age.

Analysis of genome-wide data sets from mice, dogs, and humans indicated that the researchers' hypothesis had merit: numerous CpGs in the rDNA exhibited signs of increased methylation—a result of aging. To further test the clock, they studied data from 14-week-old mice that responded to calorie restriction, a known intervention that promotes longevity. The mice that were placed on a calorie-restricted regimen showed significant reductions in rDNA methylation at CpG sites compared with mice that did not have their caloric intake restricted. Moreover, calorie-restricted mice showed rDNA age that was younger than their chronological age.

The researchers were surprised that assessing methylation in a small segment of the mammalian genome yielded clocks as accurate as clocks built from hundreds of thousands of sites along the genome. They noted that their novel approach could prove faster and more cost effective at determining biological and chronological age than current methods of surveying the dispersed sites in the genome. The findings underscore the fundamental role of rDNA in aging and highlight its potential to serve as a widely applicable predictor of individual age that can be calibrated for all mammalian species.

Importantly, the clocks respond to interventions, which could allow scientists to study how biological age responds to environmental exposures and lifestyle choices. Being able to ascertain an accurate biological age can give a person an indication of how much better or worse he or she is doing relative to the general population and could potentially help monitor whether someone is at heightened risk of death or a certain disease.

https://medicalxpres...ing-clocks.html


  • Informative x 2

#143 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 14 February 2019 - 08:24 PM

Uncovering a 'smoking gun' of biological aging clocks

A newly discovered ribosomal DNA (rDNA) clock can be used to accurately determine an individual's chronological and biological age, according to research led by Harvard T.H. Chan School of Public Health...

 

 

Thanks. The original publication is here:

https://genome.cshlp...7/gr.241745.118



sponsored ad

  • Advert
Advertisements help to support the work of this non-profit organisation. [] To go ad-free join as a Member.

#144 albedo

  • Guest
  • 1,323 posts
  • 402
  • Location:Europe
  • NO

Posted 15 February 2019 - 05:35 PM

Thanks. The original publication is here:

https://genome.cshlp...7/gr.241745.118

 

The paper is quite complex if you wish to give a try. Here is a bit of deciphering by Steve Hill of LEAF:

https://www.leafscie...biological-age/


  • Informative x 1





Also tagged with one or more of these keywords: anti-aging

2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users