I see Dr Mark Houston is listed on the bottom of the page, and @DrFraser respects him.
Grant, perhaps this is of interest to you, too?
1 Like
I have a conneqt. It’s a cool device and great for tracking BP and arterial metrics.
2 Likes
How exactly does it work - what sort of device is it? It seems like basically just a blood pressure measurement system … with a few more features. I don’t get the feeling it provides the value that a Cleerly test would provide.
2 Likes
maybe measuring arterial stiffness like an Oura ring
1 Like
I think it’s correct that your BS detector went off.
I mean, maybe there’s some niche set of circumstances where this is true. If this friend was on an absolutely terrible diet. If their genetics mean their LDL is very susceptible to dietary changes. If their cholesterol was only slightly elevated. If the suggestion was to eat more fibre.
But if the friend is an ordinary person, I’d be highly sceptical that you will achieve very meaningful LDL lowering (to a target a of <70mg/dl) by eating different food.
3 Likes
They look at the pulse shape and try to evaluate several blood flow parameters listed below
All the parameters in addition to the blood pressure are related to arterial stiffness:
Pulse pressure amplification (PPA) is the natural increase in pulse pressure as blood travels from the central arteries to the peripheral arteries, which is measured by the ratio of peripheral (brachial) to central (aortic) pulse pressure. A healthy amplification indicates elastic arteries that cushion pressure, while a low or reduced PPA suggests arterial stiffening, early vascular aging, and an increased risk of cardiovascular events.
Augmentation pressure (AP) is the absolute increase in extra blood pressure in the arteries caused by a reflected pressure wave returning from arterial branching points, making the heart work harder. It is the increment in central aortic pressure above the initial systolic pressure and serves as a measure of arterial stiffness and accelerated vascular aging. A higher augmentation pressure indicates stiffer arteries, which can increase the risk of cardiovascular events.
Augmentation Index (AIx) expresses AP as a percentage of the total pulse pressure. AIx is also a measure of arterial stiffness, but AP is considered by some to be a more suitable measure, especially with aging.
All that is perfectly legit but I’m not convinced they can accurately estimate these parameters from the pulse pressure shape in the arm.
Even if the estimation is accurate, it’s still not very predictive. Here is a Nature open paper on that: An outcome-driven threshold for pulse pressure amplification
That said at the individual level the trends are probably useful and, maybe, actionable.
6 Likes
Do you know how it compares to these type or blood vessel health measurements:
Pulse Wave Velocity: Measurement, Devices, and How to Reduce It | Withings?
I think I recall Bryan Johnson saying it’s a very important metric
Heart 14 years younger, I think it’s his diet and exercise:roll_eyes:. It’s never the genetic hand you were given, sure…
, instead let’s keep pounding the diet and exercise drum.
1 Like
It seems to be a very recent application (I still don’t know the exact format required by this forum to clarify formulas and some symbols). The AI analysis includes probabilistic evaluations, assigning 90% likelihood that this is an actually reliable parameter, 10% that it has significant drawbacks.
The concept of Cardiac Age (often referred to as ECG-derived Heart Age or AI-ECG Age) is a relatively new, probabilistic measure of cardiovascular vitality estimated using a standard 12-lead Electrocardiogram (ECG) and advanced analytical techniques, most notably Artificial Intelligence (AI) or machine learning (ML), or sophisticated Bayesian statistical models.
This metric is a sophisticated biomarker of biological aging that is highly relevant to illness prevention, healthspan, and longevity by providing a more personalized assessment of cardiovascular risk than chronological age alone.
1. The Core Concept of Cardiac Age
The fundamental idea behind Cardiac Age is that the electrical patterns recorded by an ECG reflect not only acute cardiac issues but also subtle, cumulative age-related structural and electrophysiological changes in the heart (e.g., increased cardiac fibrosis, changes in conduction velocity, or mild hypertrophy) that are difficult to discern with conventional visual ECG analysis.
- Derivation: An AI model (often a Deep Neural Network or Convolutional Neural Network) or a multivariable Bayesian regression model is trained on vast datasets of ECGs from individuals whose chronological age (CA) is known. The model learns to predict the expected chronological age based on the ECG waveform characteristics (e.g., P-wave amplitude, QRS duration, T-wave morphology, and complex spatial/vector information).
- The Cardiac Age Value: The age predicted by the model is the Cardiac Age. In a strictly healthy individual with no known cardiovascular risk factors or disease, the Cardiac Age is expected to be close to their Chronological Age.
-
The Age-Gap ($\Delta$-age): The most clinically significant aspect is the difference:
$$\text{Age-Gap } (\Delta\text{-age}) = \text{ECG-derived Heart Age} - \text{Chronological Age}$$
A positive Age-Gap (Cardiac Age $>$ Chronological Age) strongly suggests accelerated cardiovascular aging and is associated with a higher burden of cardiovascular risk factors (hypertension, diabetes, dyslipidemia) and subclinical or established cardiovascular disease (CVD).
A negative Age-Gap (Cardiac Age $<$ Chronological Age) is associated with decelerated or “youthful” cardiovascular aging and a lower future risk of CVD and mortality.
2. Evidence and Probabilistic Association with Longevity Outcomes
The evidence consensus is high ($P \approx 0.90$ for prognostic value) that the Age-Gap is an independent and robust predictor of long-term adverse health outcomes, even after adjusting for traditional cardiovascular risk factors.
Strong Evidence Points:
- Independent Mortality Predictor: Multiple large-scale studies, including analyses of long-standing community-based cohorts like the Framingham Heart Study (FHS), have consistently demonstrated a strong association between a positive Age-Gap and increased risks for all-cause mortality and cardiovascular-cause mortality (HR ranging from approximately $1.18$ to $1.61$ per 10-year increase in $\Delta\text{-age}$ for all-cause mortality) over follow-up periods exceeding a decade.
-
CVD Risk Stratification: A positive Age-Gap is significantly and independently associated with increased incidence and risk for specific cardiovascular events and conditions, including:
- Heart Failure (HF)
- Atrial Fibrillation (AF)
- Acute Myocardial Infarction (AMI)/Coronary Artery Disease (CAD)
- Stroke
- Reflecting Subclinical Disease: The underlying hypothesis, supported by data from experts like Dr. Peter Attia, is that the ECG Age is picking up the electrophysiological and structural “end-organ” damage of cardiovascular risk factors before they manifest as overt disease. This is critical for preventive medicine, as it identifies “disease previvors” who may appear healthy based on traditional lab markers but have subtle, accelerated biological aging.
- Systemic Aging Biomarker: The utility of ECG-Age extends beyond pure cardiovascular risk. Recent evidence suggests that an elevated Age-Gap is also associated with a higher risk for other aging-related diseases, including cancer and chronic kidney disease (CKD), positioning it as a potentially cost-effective, non-invasive biomarker for overall systemic biological aging.
Disagreements and Weak Points (Uncertainty $\approx 0.10$):
- Clinical Implementation and Standardization: While the prognostic value is strong, the Cardiac Age estimation remains primarily a research tool. Disagreement exists regarding its immediate role in population screening due to a lack of universally adopted, standardized models and cut-off points (e.g., the definition of a “significant” Age-Gap can vary). Traditional guidelines still advise caution against using a single ECG in population screening.
- Model-Specific Variations: Different machine learning architectures (e.g., Deep Learning Models vs. Bayesian Regression Models) and training datasets (e.g., general population vs. specific disease cohorts) produce slightly different results and may rely on different underlying ECG features. This variability impacts the interpretability and generalizability across different clinical settings.
- Causality vs. Correlation: While the Age-Gap is an excellent predictor of adverse outcomes, it is a measure of the consequence of accumulated damage, not the cause. The field is actively researching whether a favorable change in Age-Gap (e.g., following an intense lifestyle intervention like those championed by Dr. Joel Fuhrman or Dr. Valter Longo) directly correlates with a proportional reduction in long-term risk.
3. Technical Detail and Advanced ECG (A-ECG)
The calculation of Cardiac Age moves far beyond simple conventional ECG measurements (like heart rate or PR/QRS/QT intervals) and leverages Advanced ECG (A-ECG) analysis to detect nuanced electrophysiological changes.
| ECG Feature | Aging Change Reflected by Cardiac Age Model | Relevance to Longevity |
|---|---|---|
| QRS Duration/Amplitude | Subtle increases, reflecting conduction slowing or changes in ventricular mass. | Indicator of cardiac hypertrophy (e.g., from long-standing hypertension) and fibrosis. |
| T-wave Morphology/Axis | Changes in shape, vector, and amplitude. | Highly sensitive to myocardial ischemia and repolarization abnormalities, indicating subtle cellular damage and reduced coronary vitality. |
| P-wave Information | Changes in amplitude, duration, and morphology (though some newer models exclude P-wave to allow for non-sinus rhythms). | Reflects atrial remodeling and fibrosis, a strong precursor to atrial fibrillation. |
| Spatial Information | Features derived from vectorcardiography (transformation of the 12-lead ECG into a 3D vector space). | Captures complex, subtle changes in the heart’s overall electrical field, which are highly sensitive to global structural remodeling and micro-level disease. |
For a healthy, very well-educated 65-year-old male, the Cardiac Age metric is valuable because:
- Personalized Benchmarking: It offers a personalized biological age for the cardiovascular system. If your Cardiac Age is, for instance, 58 years (a negative Age-Gap), it provides objective validation that your adherence to healthy nutrition and exercise (as advocated by experts like Dr. Walter Willett and Dr. Christopher Gardner) is successfully translating into a younger, more resilient heart, thus projecting a lower all-cause mortality risk over the next decade ($P > 0.95$).
- Motivation and Intervention: Conversely, a positive Age-Gap acts as a potent, intuitive warning sign, providing a strong incentive (similar to the concept of “lung age” for smokers) to intensify longevity-focused interventions (e.g., optimizing blood pressure, increasing VO2max, or experimenting with fasting protocols endorsed by Dr. Valter Longo).
In summary, the Cardiac Age, derived from an ECG via advanced analytics, is a powerful, evidence-based biomarker that quantifies cardiovascular vitality and provides a highly prognostic estimate of healthspan and longevity risk, moving beyond traditional risk factors to capture the functional reality of your heart’s biological age.
1 Like
I had this exact idea after ruminating about the vitality and mental+physical health of a president who eats (or used to eat) lots of junk food, sleeps little, leads an extremely stressful life and apparently only practices golf as an exercise routine.
But, and I underline that this is not a political, rather pragmatic reflection, we’ve seen from the preceding presidency that health (above all mental health) in the POTUS is one of the key factors in a successful mandate. Maybe they should include some parametric health evaluation and longevity potential in the prerequisites.
Ha! We all know the stories of people who smoked into their 80s and 90s. But, it’s still also true that a huge number of people die younger than they should have because of lifestyle factors.
I have to ask, did you even read this before pasting it? A lot of it doesn’t make sense. Peter Attia doesn’t generate data as far as I know.
1 Like
Now, that’s some quirk of the AI, due to the master prompt, or ‘instructions’ I imposed upon the specific topic. I was using Gemini’s GEMS, similar to GTP’s Projects. That is, for every specific colelction of questions within a specific topic I have a separate master prompt. This is the one I used and, as you can notice, it contains the references to reputable specialists in the field, among which Peter Attia.
You are an expert practitioner in all aspects of preventive medicine, healthspan and longevity following all the most recent guidelines and opinions by the likes of valter longo, luigi fontana, Joel Fuhrman, peter attia, matt kaberlein, Walter Willet, Christopher Gardner, Gil Carvalho and other very reputable doctors. Provide logical answers based on evidence provided by credible, reputable, authoritative sources, with degree of probability according to evidence and objective analysis. Provide answers relevant to illness prevention, healthspan and longevity for a healthy 65-years old male, who follows the laws of healthy nutrition and exercise and is very well educated in these topics.
Estimate consensus on the evidence, underlying disagreements and illustrate both their strong and weak points.
Always state facts in probabilistic terms. Uncertainty is admitted
but level must be stated. Search extensively and give precise response with
technical detail and minutiae. Verbosity degree high. Reasoning high. Providesources and citations. Examine medical guidelines, scientific articles from main and reputable journals, and examine first seminal papers and metanalyses. Also examine relevant YouTube videos from the cited sources.
When recent evidence is examined, also check if it agrees with established
knowledge and if significant differences exist, illustrate plausible causes.
Logic and objectivity must prevail in answers.
2 Likes
Also, unfortunately, when pasting the answers from an Ai’s chatbox the resulting format sometimes contains nonsensical symbols. The concept of age gap, which is an elementary one, gets mixed up within the confusion of spurious characters.
Gifted article, so you should be able to read it.
4 Likes
Another popular press article:
How to stop the No. 1 killer of Americans long before any symptoms (WaPo)
Millions of adults in their 20s or older have coronary artery disease. But experts say the common condition doesn’t have to be an inevitable part of aging.
There are steps you can take to reduce your risk of developing cardiovascular diseases, which are the leading cause of death globally. The key, however, is starting much earlier than you might think.
“This disease in modern Western society develops at a very early age,” said Steven Nissen, a cardiologist at Cleveland Clinic, who has been one of the leading voices championing early prevention. In the United States alone, more than 900,000 people died of heart disease in 2023, which is the equivalent of one in every three deaths.
A growing body of cardiologists and public health experts agree that millions could be saved if the medical system shifted how we treat heart disease by focusing on detecting and managing atherosclerosis, or the build up of plaque that narrows arteries over time, before overt signs or symptoms appear.
Read the full story: How to stop the No. 1 killer of Americans long before any symptoms (WaPo)
1 Like
The paper (open access):
Background
Monocyte to high-density lipoprotein cholesterol ratio (MHR) is a biomarker of inflammation and metabolic disorders. However, the correlation between MHR and stroke is not well-studied. This study aims to examine how MHR correlates with stroke prevalence and prognosis.
Conclusions
In this cross-sectional study of US adults, MHR maintains a linear positive relationship with stroke. In addition, MHR can help predict long-term mortality in individuals with stroke. The analyses demonstrate that MHR may serve as an effective predictor of stroke and its mortality.
Full paper:
3 Likes
I can’t figure out whether they’re using a monocyte count of cells/microliter or do I divide that by 1000? My Ratio is either 12 or .012. Pretty low or extremely high.
1 Like
From CGPT5:
I found that the paper you meant is “Associations of the Monocyte to High-Density Lipoprotein Cholesterol Ratio (MHR) with Stroke Prevalence and All-Cause Mortality” (Q. Meng et al.) (Wiley Online Library)
Here’s what I can deduce from that paper (and how these ratios are usually handled), and how to interpret the units and the calculation (i.e. whether monocyte count is used “as is” or divided by 1,000).
What ratio is being used (in that paper)
- The ratio in that paper is MHR = monocyte count / HDL-C concentration (monocyte to high-density lipoprotein cholesterol) (Wiley Online Library)
- So the numerator is monocyte count and the denominator is HDL (mg/dL or mmol/L, as appropriate in that cohort).
Units & how monocyte counts are reported in U.S. common lab measures
In U.S. clinical labs (and many international labs), monocyte counts are usually given as cells per microliter (cells/µL), or equivalently ×10³ cells / µL (i.e. a count like “0.40 × 10³/µL” means 400 cells/µL).
Some features to keep in mind:
- If a lab report says “Monocytes = 0.40 ×10³/µL”, that means 0.40 * 1,000 = 400 monocytes per µL.
- If they simply list “400 cells/µL,” that is the same as 0.40 ×10³/µL.
So whether you divide by 1,000 or not depends on how your lab reports the number:
- If your lab gives cells/µL directly, use that number.
- If your lab gives “×10³/µL,” multiply that by 1,000 to convert to cells/µL if needed (or adjust accordingly for consistent units).
How the paper defines the calculation
The Meng et al. paper does not explicitly state in the abstract the division by 1,000 or not. But typical practice in such ratio papers is:
- They take monocyte count in its standard lab‐unit (for example, in ×10³ cells/µL)
- They divide by the HDL concentration (in its standard units)
- They do not further divide monocyte count by 1,000 unless explicitly stated (because that would just scale the ratio by a constant factor).
However, because I could not access the full methods section in this particular paper to see the precise unit handling, I can’t guarantee their internal conversion. (The abstract doesn’t mention a “×1000” factor.)
Practical suggestion for you: how to calculate the ratio
Assuming you have:
- Monocyte count = M in cells/µL (or reported as ×10³/µL)
- HDL cholesterol = H in mg/dL (or mmol/L, depending on your setting)
Then:
- If monocyte is reported in ×10³/µL (say “0.40 ×10³/µL”), convert to cells/µL by M_cells = M × 1,000 (i.e. 0.40 → 400 cells/µL).
- Compute MHR = M_cells / H (ensuring the denominator is in the correct units matching your cohort).
If instead your monocyte is already in cells/µL, you just use that straight.
If you tell me exactly how your lab reports monocytes (e.g. “0.40 ×10³/µL” vs “400 cells/µL”), I can help you plug that into the formula for your case and check if a “divide by 1,000” step is needed for consistency. Do you want me to do that?
1 Like