I work at a biotech that does intersecting research (predicting disease risk). We obsess about temporal variation at various time scales and the effects are quite real. We use multiple data planes (multi-omics) to evaluate temporal effects across diverse biology. Our goal is to ensure that our long-term predictions are minimally impacted by short-term fluctuations.
Fair enough, could I prod a bit about this section in particular? ("We use multiple data planes (multi-omics) to evaluate temporal effects across diverse biology."
So you're a high-tech epidemiology company providing disease risk assessments based on your predictive models. These predictive models you say are using multiple data planes to evaluate temporal effects across diverse biology.
My questiosn are, are data planes (multi-omics) the same as just data sources. Naturally conducting epidemiology necessitates taking in data from many sources, where does the term multi-omics come into play here? Presumably some of your data sources are genetics so I could see the "omics" part coming from that field, but it's an interesting term and I'd like to know what you mean by it?
I think I can relate to this, even despite mine is just purely software startup (but it's low-level software/firmware for off-the-shelf devices, which can only run commercial closed ones at the moment). Too much attention before our product is mature enough would be easily able to kill that. We won't be able to handle whatever could arise from that with our several-person team.
But that's me. Mind sharing your reasons? Of course could be obscured as much as you need.
In this context, stealth helps with trust and communication. Although that might seem paradoxical, we are initially working with a small number of members reached by word of mouth.
We walk members through our data usage policies in person and ask questions to ensure we understand both their health and privacy concerns. Without long-term trustworthiness on our part, we won't get the longitudinal data that we need to ensure that we continue building the best in class predictive models.
There are other reasons, but that's the biggest one.
One of the "dirty secrets" in healthcare is similar issues with BP.
Things like posture, white-coat syndrome, having to pee, cuff placement, cuff size, arm position, muscularity, arm diameter, time sitting, and many other variables have a massive influence on blood pressure.
BP is kind of like weight - it should be taken at regular intervals during the day, at the same times, and averaged out over time to look at trends. "Snapshot" BP readings are most useful for things like hypertensive crisis, not ongoing BP management.
Unfortunately, we see a lot of people put on BP meds where other interventions may be more appropriate based on bad BP measurements.
I think there are a lot of things in healthcare like this, where our models are too simplistic and result in flawed understanding and consequently ineffective treatments.
I have really bad white-coat hypertension. Like 150-90 in the drs. office. So every time I go in they send me home with a chart and I have to dutifully record my blood pressure a couple of times a day for 2 weeks and send it back to them, and when measured at home it's totally fine.
Since I have the cuff I check it every so often to make sure it's still fine and it is, but the moment I get in a drs. office boom it spikes. This is even true when I'm there for someone else like my son or my wife. I can feel that my blood pressure is high.
I have absolutely no explanation for why except that I had a major surgery when I was 5 and there must be some suppressed trauma from that that comes out physically when I'm in a doctor's office.
To my mind it’s a very natural reaction. We rarely are in situations, like a doctor’s office, where out of the blue you may be told you are dying, etc.
I realized this a couple years ago when I moved and started getting my blood tests done at a new lab that was nearby.
My entire life I’ve had perfect blood pressure, but suddenly this lab thinks I have hypertension. After 3-4 visits and checking my own BP during the day, I realized that my blood pressure is just higher in the morning - I never set appointments in the morning previously because I lived too far from a lab to make morning appointments convenient.
What makes them think you have hypertension with purely blood laboratory tests? Electrolyte levels ratios? They aren't changed in all types of hypertension (won't be changed in most, probably), but that's just my wild guess.
Is it really secret though? It's the first thing you learn when you google "how to take bp". Like I'm not a medical person but when I was pregnant, I was asked to check BP and I asked the doctor "should I like average 3 readings or something?" and she said "just take one, don't stress it".
This is well known re thyroid medication as well. Also re weight, progesterone, a bunch of things.
It's not a secret. Most of my immediate family is or was in healthcare and nobody is getting put on BP meds because of a single errant reading. They're getting put on meds because of persistent, uncontrolled elevated BPs and with non-pharmaceutical interventions not addressing it. Everyone, especially physicians, know there's a list of 20-30 things that can give you temporarily elevated BP that won't respond to BP meds.
Yeah, that's been my experience. Everytime I see my primary my BP is elevated in the office and they send me home with a log and I have to take my BP multiple times a day for 2 weeks and at home it's fine, so I've never been put on meds for it even though it's very high in the office.
I suppose "well known" is relative, it may surprise you how little agency some people have over their health.
>It's the first thing you learn when you google "how to take bp".
The populations that I frequently deal with are not in a life/financial/cognitive place where this is a thing they would think of doing, or be capable of doing.
Take a look at this [1] about over treatment of mild hypertension. It's not a new problem, and potentially affects millions [2].
There can be enormous pressure on physicians to treat certain conditions from a population-health standpoint with medications if they fit certain clinical criteria.
The recent (2017) change in standards for hypertension [3] have been controversial. It put millions of people into the category of hypertension that weren't previously. Coupled with certain clinical criteria that recommend medication prescription even for mild hypertension, it's concerning that there may be an overprescription issue.
Many busy clinics don't have time to have a patient sit for the recommended 5 minutes of rest before taking BP (though I've seen this improving). Many medical assistants are poorly trained in BP measurement, and are overworked - I've seen many cases where a harried MA will do things like sit a patient down, put a cuff on them over-the-sleeve, as the patient slouches in a chair, with the arm in an incorrect position, a poorly fitted cuff, and just write down the number the machine spits out.
Of course, a good PCP will double check this and look for a trend. Not all PCP's are good and go to this effort.
(I am not a physician, I'm a tech nerd with a clinical background, but I've worked in healthcare for decades)
I've also experienced it personally. I went to the doc once about ten years ago for a wellness check. Based on the in-office BP reading and labs, the NP prescribed me five different medications, including BP meds, statins and others.
I decided to focus on lifestyle changes instead and that's worked well for me, but that's not an option for everyone.
Yes, blood pressure measurement is a real mess. Besides the factors you listed, I have found that temperature has a surprisingly large impact. If your hands are cold, then that causes peripheral vasoconstriction and raises BP.
Question, since you seem to know a lot about this: is high blood pressure meant to be a description of a persistent condition (I.e. took a reading at 5, 10, 20 minute intervals at rest and stayed elevated) or is it meant to be a description of something else (after standing up, or after walking, etc etc)?
Generally, it's referring to a chronic condition, meaning that your blood pressure is elevated as a trend over time.
Everyone goes through huge peaks and troughs with BP - at the gym, sleeping, etc... will have massive changes in BP.
There are certain conditions where "instant" BP can become a problem, both high and low. Orthostatic hypotension, for example, is a condition where you can get dizzy or black out from changing position too quickly.
There's also a condition known as labile hypertension that can cause random high spikes in BP.
This should be basic knowledge for any provider. I don't really agree that it is a model that is too simplistic. It is a tool that has considerations for accurate use based on physics and physiology.
The potential for systematic bias here is also quite concerning. Imagine your experimental group tends to come in the morning (maybe because they're on-site) but your control group is scheduled to come in the rest of the day...
This problem is VERY real and happens in surprising ways. My biotech has a dataset that was collected over several years and we found this pattern of older people having more Vitamin D in their blood than younger people, which is the opposite of the published literature. It turns out that the sample collection was initially year round but switched to summer only and everyone in the cohort was older, by definition, in the later collection.
Source? Because I know people working on it as a target measure at a national lab in India as well as at Pfizer (in America) and Novartis (in Switzerland).
Epigenetic alterations such as DNA methylation are only one hallmark of aging, and not the most useful one in most cases. Interventions to reduce those epigenetic alternations have generally not proven effective in extending lifespans or improving health outcomes.
> Epigenetic alterations such as DNA methylation are only one hallmark of aging, and not the most useful one in most cases. Interventions to reduce those epigenetic alternations have generally not proven effective in extending lifespans or improving health outcomes
Thank you. And I totally agree. It reminds me of telomere craze in the early noughties.
It not being a panacea, however, is different from it being useless. Marking it as pure marketing makes it sound like it's cosmetics. It's not. The first research on chronological age prediction dates from 2011; we're still in the basic research phase. (Which the article mentioned is. It's not marketing any treatment.) To the degree it's being turned into a scam it's in supplements, not bespoke medicine.
> then we just disagree on this and history will see who is right
That’s not how science works.
Epigenetic age prediction is an area of active research for general diagnostic and treatment vectors [1]. There are peer-reviewed studies on the stuff. You claimed “really a marketing tool for bespoke medicine”. That appears to be false.
That doesn't really contradict the post you're responding to, does it? (they say it is not that useful for bespoke medicine, you say that people are using it for research purposes, both can be true?)
> they say it is not that useful for bespoke medicine, you say that people are using it for research purposes, both can be true?
They said it is “really a marketing for bespoke medicine.” That implies only (or mostly) useful to bespoke medicine as a marketing tool. Demonstrating any significant other utility falsifies that statement.
Still both can be true. I'd say it's even a common thing in medicine. Researchers may be researching something real with well-defined range of effect/application, meanwhile many a scoundrel will pick up the same term to market their bullshit treatments or lifestyle advice, and when interrogated, they'll actually point at some of the published studies of the real thing, hoping people won't understand the difference.
I mean epigenetic clocks are extremely unreliable and have enormous error bars. It is not surprising that they are easily influenced by additional variables.
Epigenetic clocks are one of the (if not the) most accurate ways to estimate someone's real chronological age using molecular level data. So not sure what you mean by "extremely unreliable".
I disagree, because "real age" is defined in terms of the epigenetic clock. We don't have enough time observing them with humans to connect them to real age, at least as far as I have ever seen. If there's some major paper I have missed, and my other sources have missed, I'd love to see it.
If you were born 50 years ago today, you're 50 years old. Age is a concrete, mathematical thing. That your cells appear to be younger or older to your actual (chronological) age is irrelevant. You don't get to say your "real age" is 45 because your cells are slightly better than they "should be."
“I am 50 years old” is a statement of fact like you refer.
“Real age” is essentially a term of art used to describe your relative health - a single number to indicate the cumulative impact of your lifestyle, genes, etc.
No one says “I have a real age of 50” to the question “how old are you?”
Thanks. Moreover nobody would use an epigenetic clock to refer to chronological age, because it's inferior in every way for that purpose.
The purported use of an epigenetic clock in humans is that it predicts either aging effects of remaining life span in some sense; neither of these have been proven to be biologically true because in the first we don't necessarily have the definition of aging, and certainly don't have it down to a single axis, at best it is understood along many different dimensions, many of which are not captured by epigenetic measures.
And second sense, predicting remaining life span, is completely unproven. Even more speculative is whether measures which change a person's epigenetic age will result in changes in remaining lifespan.
There's a chance that epigenetic age becomes useful for something, but it's just a chance. It's not established science.
They probably are the most accurate way, but that still doesn't make them good. If there haven't been any recent breakthroughs I missed, then we are talking about getting it right +/- 4 years under ideal conditions. Repeating these test on the same subjects can result in fluctuations of 8 to 12 years [1].
Additionally, there are a lot of biotech/junk studies claiming some new intervention reduces/reverses biologic age using these epigenetic clocks as evidence. Making me very wary of the whole thing, though unfairly to those actually doing sound studies I admit.
I think we mostly agree. If you are interested, there are developments that reduce the fluctuations of clock predictions - like principal component based clocks [1], where the authors claim average difference between replicate measurements of 1.5 years.
And the study you are citing (8 to 12 year errors) - seems to report maximum errors, not averages. These might be caused by poor quality samples, so it's unfair to report it like that. Also the publication under discussion (epigenetic age oscillates) would answer where the fluctuations observed by the article you linked to come from - part of them are of course technical due to measurement error, but part of them are influenced by differences in sample collection times.
I also share your skepticism about intervention research and using "biological clocks" to measure how healthy you are. Curiously chronological clocks seem to not be easily affected by interventions. And I would guess this is why most people doing epigenetic aging for sport use biological clocks (like DunedinPACE). But then the question - if you are younger according to "biological age", but we still can measure your true chronological age accurately - are you actually younger in a meaningful sense.
> My temp will be 98.4 while seated but if I stand up and check it, it drops down to 97.9
Body temperature measurements from outside the body are impacted by changes in blood flow, which will happen when you stand up.
Your core body temperature isn't fluctuating, the temperature at the measurement point is changing slightly because you've changed your blood flow.
We're constantly losing heat from our bodies to the environment. You could probably get similar measurements if you heavily insulated the entire area around the thermometer and let measurements stabilized.
In other words: Don't read too much into this experiment.
Maybe its just not a good enough measure for any form of age? Some folks get obsessed with numbers without seeing bigger picture and focusing on actually important things in life (just general observation, please don't take any of this personally)
How are you taking your temperature? Infrared sensor on the forehead? Infrared only measures surface temperatures and can give false readings of temperature based on how the body "radiator" system is working.
The fluctuations aren't that bad. Now, I don't trust the infrared sensors myself, because I can't see the "local gradient" of what they're measuring, nor how it changes momentarily (why does it always take a second or two to measure?), and there certainly are areas on the face that vary by whole degree centigrade in temperature.
How do I know? Because I got so annoyed at the IR thermometers that I recently got myself one of those USB-C thermal cameras (UTi721M, specifically). It's like a 256x192 array of IR thermometers, measuring continuously, so you actually see what's going on. Lots of interesting things you can learn that way, for example that cheeks, chin and nose can easily be 1-2℃ cooler than the forehead, or that the neck/throat shows about the same temperature that the forehead, but seems more stable. And, it's much faster to take temp of everyone at home in one go. And then you also can learn a lot about your own environment, too. I highly recommend this to anyone.
That said, I found in-ear IR most reliable, and use it as a baseline for health checkups. You do however need to watch out for insertion - if you put the head in the ear too fast and measure immediately, you can get a result that's a degree or two centigrade above correct, which I imagine is because of momentary friction heating.
So you're a high-tech epidemiology company providing disease risk assessments based on your predictive models. These predictive models you say are using multiple data planes to evaluate temporal effects across diverse biology.
My questiosn are, are data planes (multi-omics) the same as just data sources. Naturally conducting epidemiology necessitates taking in data from many sources, where does the term multi-omics come into play here? Presumably some of your data sources are genetics so I could see the "omics" part coming from that field, but it's an interesting term and I'd like to know what you mean by it?
But that's me. Mind sharing your reasons? Of course could be obscured as much as you need.
We walk members through our data usage policies in person and ask questions to ensure we understand both their health and privacy concerns. Without long-term trustworthiness on our part, we won't get the longitudinal data that we need to ensure that we continue building the best in class predictive models.
There are other reasons, but that's the biggest one.
Things like posture, white-coat syndrome, having to pee, cuff placement, cuff size, arm position, muscularity, arm diameter, time sitting, and many other variables have a massive influence on blood pressure.
BP is kind of like weight - it should be taken at regular intervals during the day, at the same times, and averaged out over time to look at trends. "Snapshot" BP readings are most useful for things like hypertensive crisis, not ongoing BP management.
Unfortunately, we see a lot of people put on BP meds where other interventions may be more appropriate based on bad BP measurements.
I think there are a lot of things in healthcare like this, where our models are too simplistic and result in flawed understanding and consequently ineffective treatments.
Since I have the cuff I check it every so often to make sure it's still fine and it is, but the moment I get in a drs. office boom it spikes. This is even true when I'm there for someone else like my son or my wife. I can feel that my blood pressure is high.
I have absolutely no explanation for why except that I had a major surgery when I was 5 and there must be some suppressed trauma from that that comes out physically when I'm in a doctor's office.
My entire life I’ve had perfect blood pressure, but suddenly this lab thinks I have hypertension. After 3-4 visits and checking my own BP during the day, I realized that my blood pressure is just higher in the morning - I never set appointments in the morning previously because I lived too far from a lab to make morning appointments convenient.
This is well known re thyroid medication as well. Also re weight, progesterone, a bunch of things.
I'm glad you've had good experiences and relationships with what sound like quality healthcare professionals.
Others have had different experiences.
>It's the first thing you learn when you google "how to take bp".
The populations that I frequently deal with are not in a life/financial/cognitive place where this is a thing they would think of doing, or be capable of doing.
Take a look at this [1] about over treatment of mild hypertension. It's not a new problem, and potentially affects millions [2].
There can be enormous pressure on physicians to treat certain conditions from a population-health standpoint with medications if they fit certain clinical criteria.
The recent (2017) change in standards for hypertension [3] have been controversial. It put millions of people into the category of hypertension that weren't previously. Coupled with certain clinical criteria that recommend medication prescription even for mild hypertension, it's concerning that there may be an overprescription issue.
Many busy clinics don't have time to have a patient sit for the recommended 5 minutes of rest before taking BP (though I've seen this improving). Many medical assistants are poorly trained in BP measurement, and are overworked - I've seen many cases where a harried MA will do things like sit a patient down, put a cuff on them over-the-sleeve, as the patient slouches in a chair, with the arm in an incorrect position, a poorly fitted cuff, and just write down the number the machine spits out.
Of course, a good PCP will double check this and look for a trend. Not all PCP's are good and go to this effort.
(I am not a physician, I'm a tech nerd with a clinical background, but I've worked in healthcare for decades)
I've also experienced it personally. I went to the doc once about ten years ago for a wellness check. Based on the in-office BP reading and labs, the NP prescribed me five different medications, including BP meds, statins and others.
I decided to focus on lifestyle changes instead and that's worked well for me, but that's not an option for everyone.
[1] https://www.youtube.com/watch?v=cZCuJheVJA0
[2] https://www.umassmed.edu/news/news-archives/2014/11/umms-fam...
[3] https://www.health.harvard.edu/heart-health/reading-the-new-...
[0] https://www.frontiersin.org/articles/10.3389/fcvm.2019.00040...
[1] https://www.cnet.com/health/medical/i-have-high-hopes-for-th...
Everyone goes through huge peaks and troughs with BP - at the gym, sleeping, etc... will have massive changes in BP.
There are certain conditions where "instant" BP can become a problem, both high and low. Orthostatic hypotension, for example, is a condition where you can get dizzy or black out from changing position too quickly.
There's also a condition known as labile hypertension that can cause random high spikes in BP.
It sounds like the most immediate takeaway is that anyone analyzing this stuff needs to control for when measure individuals.
[1]: https://www.nature.com/articles/s41467-023-42588-6
Source? Because I know people working on it as a target measure at a national lab in India as well as at Pfizer (in America) and Novartis (in Switzerland).
https://doi.org/10.1016/j.cell.2022.11.001
Epigenetic alterations such as DNA methylation are only one hallmark of aging, and not the most useful one in most cases. Interventions to reduce those epigenetic alternations have generally not proven effective in extending lifespans or improving health outcomes.
Thank you. And I totally agree. It reminds me of telomere craze in the early noughties.
It not being a panacea, however, is different from it being useless. Marking it as pure marketing makes it sound like it's cosmetics. It's not. The first research on chronological age prediction dates from 2011; we're still in the basic research phase. (Which the article mentioned is. It's not marketing any treatment.) To the degree it's being turned into a scam it's in supplements, not bespoke medicine.
That’s not how science works.
Epigenetic age prediction is an area of active research for general diagnostic and treatment vectors [1]. There are peer-reviewed studies on the stuff. You claimed “really a marketing tool for bespoke medicine”. That appears to be false.
[1] https://pubmed.ncbi.nlm.nih.gov/34415665/
They said it is “really a marketing for bespoke medicine.” That implies only (or mostly) useful to bespoke medicine as a marketing tool. Demonstrating any significant other utility falsifies that statement.
If you were born 50 years ago today, you're 50 years old. Age is a concrete, mathematical thing. That your cells appear to be younger or older to your actual (chronological) age is irrelevant. You don't get to say your "real age" is 45 because your cells are slightly better than they "should be."
“I am 50 years old” is a statement of fact like you refer.
“Real age” is essentially a term of art used to describe your relative health - a single number to indicate the cumulative impact of your lifestyle, genes, etc.
No one says “I have a real age of 50” to the question “how old are you?”
The purported use of an epigenetic clock in humans is that it predicts either aging effects of remaining life span in some sense; neither of these have been proven to be biologically true because in the first we don't necessarily have the definition of aging, and certainly don't have it down to a single axis, at best it is understood along many different dimensions, many of which are not captured by epigenetic measures.
And second sense, predicting remaining life span, is completely unproven. Even more speculative is whether measures which change a person's epigenetic age will result in changes in remaining lifespan.
There's a chance that epigenetic age becomes useful for something, but it's just a chance. It's not established science.
Additionally, there are a lot of biotech/junk studies claiming some new intervention reduces/reverses biologic age using these epigenetic clocks as evidence. Making me very wary of the whole thing, though unfairly to those actually doing sound studies I admit.
[1] https://clinicalepigeneticsjournal.biomedcentral.com/article...
And the study you are citing (8 to 12 year errors) - seems to report maximum errors, not averages. These might be caused by poor quality samples, so it's unfair to report it like that. Also the publication under discussion (epigenetic age oscillates) would answer where the fluctuations observed by the article you linked to come from - part of them are of course technical due to measurement error, but part of them are influenced by differences in sample collection times.
I also share your skepticism about intervention research and using "biological clocks" to measure how healthy you are. Curiously chronological clocks seem to not be easily affected by interventions. And I would guess this is why most people doing epigenetic aging for sport use biological clocks (like DunedinPACE). But then the question - if you are younger according to "biological age", but we still can measure your true chronological age accurately - are you actually younger in a meaningful sense.
[1]: https://www.nature.com/articles/s43587-022-00248-2
Body temperature measurements from outside the body are impacted by changes in blood flow, which will happen when you stand up.
Your core body temperature isn't fluctuating, the temperature at the measurement point is changing slightly because you've changed your blood flow.
We're constantly losing heat from our bodies to the environment. You could probably get similar measurements if you heavily insulated the entire area around the thermometer and let measurements stabilized.
In other words: Don't read too much into this experiment.
How do I know? Because I got so annoyed at the IR thermometers that I recently got myself one of those USB-C thermal cameras (UTi721M, specifically). It's like a 256x192 array of IR thermometers, measuring continuously, so you actually see what's going on. Lots of interesting things you can learn that way, for example that cheeks, chin and nose can easily be 1-2℃ cooler than the forehead, or that the neck/throat shows about the same temperature that the forehead, but seems more stable. And, it's much faster to take temp of everyone at home in one go. And then you also can learn a lot about your own environment, too. I highly recommend this to anyone.
That said, I found in-ear IR most reliable, and use it as a baseline for health checkups. You do however need to watch out for insertion - if you put the head in the ear too fast and measure immediately, you can get a result that's a degree or two centigrade above correct, which I imagine is because of momentary friction heating.