Biological Age (1/3)

let's talk biological age. aging markers boost health management, refine life expectancy, and enhance well-being. yet, consensus on the very idea of aging is still in the works. here's my thread 1 of 3 on biological age markers.

here's what we do know

- age is a significant marker across the board

- people age differently

- it is biological, rather than chronological, age that actually matters in medicine

clinicians in routine checkups are actually often measuring biological age. this includes max. O2 consumption, kidney function, inflammatory markers, grip strength, sit-and-reach, soft lean mass, and more. these may be insufficient in isolation - more on that later.

by looking into biology, we can understand the roots, not leaves, of this issue. this can help us come up with better solutions for anti-aging

as -omics, especially outside of classical genetics, has grown, biologists started to notice genome instability and epigenetic signatures are at the root of some of the markers clinicians use.

additionally, the microbiome shows changing age fingerprints over time

at the forefront lies the “epigenetic clock” & favorite concept of all time: METHYLATION i.e. adding a -CH3 to DNA, acting like switches to turn genes on or off

this is vital for determining cell function, directing whether it becomes a nerve, or skin cell.

this falls under epigenetic marks

improper methylation makes cells lose identity, become dysfunctional, or enter senescence—stopping division and causing inflammation

methylation patterns shift with age, revealing an organism's biological age

to make sense of this, imagine methylation as dimmer switches for genes.

they adjust gene expression based on environmental factors

over time while linked to aging, their roles are complex—some changes may speed up aging, others might protect against it

DNA has a hallway of these switches called CpGs

these act as dimmers; adding a methyl group dims the light

clusters near gene promoters control key genes various factors adjust these switches, influencing gene expression

& experiences, chemistry and environment tweak settings

some papers call methylation the 'prophet' of aging

having researched methylation for a few years, seeing correlations like the one below for the ELOVL2 promoter is standard

IN SOME WAYS, methylation patterns require aging by definition

some genes have been shown to be especially linked to aging. these can be found on GenAge

a lot of the work in this database was as a result of epigenetic clocks. let's talk about them, and cover this more, ML, and other crucial markers we are missing in thread 2 or 3.

so in that paper above, we talked about the gene ELOVL2. turns out, in the tens of thousands of genes in every cell of the body, some were effectively used to create clocks, that somewhat help use understand biological aging

Bocklandt S. et al: early epigenetic clock using CpG sites and chronological age; 3 sites showed strong age correlations

a regression model using EDARADD and NPTX2 loci achieved a 5.2-year accuracy margin.

Hannum clock changed the game, using on 71 methylation markers and clinical parameters (gender, BMI)

produced a 3.9 yr error for the primary and 4.9 for the validation cohort.

the authors focused on blood cells but the model proved applicable to other human tissue types

in came Horvath - a pan-tissue clock

this was a game-changer, and Horvath continues to innovate (next thread). this clock studied methylation at 353 sites across multiple tissues, with high accuracy: R^2 of 0.97 w/ age in the training set and 0.96 in the test set

Horvath and Hannum clocks are still widely used, due to the ability to use for blood samples in latter, and the pan-tissue coverage in former. e.g. myDNAge, Zymo Research and FOXO Technologies all actively use the Horvath clock, and Hannum's is used in many large-scale studies

the next clock covering the initial foundations of robust aging is DNA-m PhenoAge

Unlike Horvath and Hannum clocks, DNA-m PhenoAge goes beyond chronological age to predict an individual’s physiological age based on DNA methylation patterns

unlike previous clocks, DNA-m PhenoAge incorporates multiple dimensions of health, questioning using chronological age as the baseline this gives a dynamic, holistic view, with insights into age-related diseases and potential interventions

the final clock for this thread is Horvath's GrimAge

an advanced clock developed to predict biological age and lifespan using DNA methylation patterns

it incorporates markers associated with aging-related diseases and mortality, like plasma proteins and smoking history

interestingly, adjusted GrimAge scores have been linked to brain health decline and other age-related neurological changes

neuroepigenetics requires a lot more info on the ontology of genes, but these findings show how far we have come and shed a quantitative light on MDD

there are many other markers of aging, including metabolomics, histology, collagen - Precigenetics , proteomics of the blood, and transcriptomics

additionally, clinicians aren't wrong - there are ways to adjust are molecular markers based on personalized and specific physiologies

however, covering it all will require more than one post. for now, let's just remember that getting a dynamic view of the body is difficult using the lab diagnostics needed for techniques like epigenetic clocks, posing a grand threat to society as a whole.

additionally, care is needed when interpreting any one, or a combination of, these molecular, cellular and clinical/physiological markers and calling it biological age - for which ML may be leveraged.

we are working on this at my startup, and many cutting-edge startups looking to understand and treat disease phenomena effectively, in ways that go beyond the lab. aging is certainly one of these phenomena, and as our tools get developed, i would like to encourage you to leverage what we DO know:

- good nutrition

- exercise

- sleep

- mental health

- mental engagement

...as well as, in many ways, what 'feels good' over time, are our best bets to fight our one foe - death and disease.

i hope this gave some insight into biological age, and there's more to come. :)