We use regular blood work you can screen from us or anywhere else in the world, and get the most accurate aging results. What do we mean by getting most accurate aging assessment? How do we measure the accuracy? What is our research? How does this compare with other research work? Why should you use this clock over others? Continue reading if you are curious about these questions.
What is biological aging?
Biological aging is the process of becoming weaker and more vulnerable to diseases as a person gets older. Birthday-based age (also called Chronological age ) doesn't capture the information about one's health correctly. Biological age, on the other hand, is a proxy of the current health and risk status of individual health. A healthy individual is ought to be "biologically" younger than an unhealthy individual with the same chronological age. Biological age can be measured using Machine learning models called "aging clocks," which are trained on health records such as genetic testing (i.e studying the DNA patterns inherited from the parents), epigenetic testing (i.e., studying how external factors, including lifestyle and behaviors, change the way genes function in the body), regular blood work your GP/doctor would prescribe for algorithmically identifying patterns in these data points that indicate the risk of diseases. Aging is a significant risk factor for many diseases, and aging clocks can be used to predict risks and evaluate the effectiveness of preventive measures such as medicine, exercise, diet, and other healthcare interventions.
How is aging measured?
The biological age can be measured through epigenetics, blood markers, metabolomics, microbiomics, transcriptomics and inflammation
The Epigenetic Clock using the methylation sites, also known as a Horvath Clock was built to estimate the chronological age based on 353 epigenetic markers on the DNA. Also, it was identified that people with higher epigenetic age compared to chronological age have a higher incidence of various age-related degenerative diseases and all-cause mortality. Other works model the aging of an organism using the concentrations of the metabolites in metabolome as described in one of the studies
What's new about the Healome aging clock?
Healome clock uses blood clinical markers (Lipid profile, renal profile, liver profile, standard biochemistry) as opposed to Horvath's clock which uses epigenetic markers. Epigenetic markers are good to get an estimate of biological age and that's it. it doesn't provide any other interpretable information.
The markers we look at us for measuring aging progression help us to
Correct for diseases
That is to decouple the effect of aging due to diseases and measure the internal reversible aging progression
Provide actionable interpretation
Epigenetic age talks about what gene should not be methylated (switched off/switched on) to reduce age - but no one knows how to target a specific gene to be switched on or off. However, the Healome clock gives information on what blood clinical markers be changed to result in a decrease in your age acceleration.
Blood clinical work data (Lipid profile, renal profile, liver profile, standard biochemistry) allows us to correct various diseases as we measure age acceleration. As diseases are often organ-specific, it is an important research question to understand how organs age in response to various diseases and interventions compared to full-body aging. Towards this end, Healome is developing organ age clocks for the liver, kidney, and heart. We train the machine learning model to predict the biological age acceleration by fitting a Cox proportional hazard model (using Gompertz law) to predict organ-specific morbidity.
Are there any other aging clocks that are designed this way?
PhenoAge is the closest work to this. It, however, uses only a subset of 8 markers. But it doesn't do organ-specific age prediction. Healome clock is trained on 15 times more data. For full body aging, Healome clock beats this in all the metrics, that brings us to the next question
How are clocks validated? How do you compare two clocks?
Kaplain-Meier Survival analysis
Imagine you're tracking how long a group of people can go without catching a cold after taking a new vitamin. Each person's experience is like a piece of a story in a book, and you want to see how the whole story unfolds. The Kaplan-Meier estimate is like a special calculator that helps you create a big picture from all these individual stories. It helps us understand how effective something (like a vitamin) is at preventing an event (like catching a cold) over time, by keeping track of each person's experience and combining them into an easy-to-understand picture.
In the below graph, accelerated_aging is the group(represented with a blue line) where the Healome age is at least 5 years more than their birthday-based age. The decelerated_aging group (represented with the orange line) is the one where the Healome age is at least 5 years lower than their birthday-based age.
You can see that, after around your 40 birthday, the people from the orange group have a higher chance of survival than the people in the blue group. This shows that the Healome Aging clock is predictive of body aging! But don't panic if you are in the blue group, because the good news is that Aging can be slowed down or even reversed - that's what we help you achieve!
Correlations with the chronological age of the cohort
The aging clock machine learning model is supposed to predict chronological age if the user is perfectly healthy (without exhibiting any accelerated phase of aging). So, in an ideal population cohort, after correcting for diseases, the biological age should more or less correlate well with the chronological age
Healome's blood-aging clock with a coefficient of determination (R^2) of 0.88 and Pearson Correlation (R) of 0.94 overperformed not only state-of-the-art blood clocks like Insilico Medicine (R^2=0.65,R=0.8) but also approached the accuracy of far more expensive methylation-based PhenoAge clocks (R=0.94, R^2=0.88) - So, we are as accurate as a methylation panel based epigenetic test!
Hazard Ratios
A hazard ratio of 1 means the risk of the outcome remains the same with every year passing by. A hazard ratio of greater than 1 would mean that the risk of the outcome increases through the exponent of the hazard ratio each year. A good aging clock is supposed to have a higher hazard ratio - because the aging clock is supposed to capture the increased risk of death/diseases as the biological age increases. In other words, the biological aging clock would detect an increase in biological age if and only if the risk of death/diseases has increased. For 1 year increase in Healome age, the risk of death increases by 13%. The good news is that 1 year decrease in the Healome age, the risk of death decreases by 13% !!!
Clock | Hazard Ratio |
PhenoAge | 1.03 |
Healome | 1.13 |
Does slowing down aging guarantee improved health?
Aging clock-based age acceleration is a holistic metric that can accurately summarize the overall health risks as well. The following are the correlations of various conditions and how they accelerate biological age. Note that these are only average numbers based on aggregated research, and your personal snapshot may differ widely ( the best way to assess it is to get your blood test done and start tracking )
Condition | Age acceleration per year |
Obesity | +0.70 years |
Heavy alcohol consumption | +2.50 years |
Former smoking | +0.75 years |
Current smoker | +1 year |
Diabetic on treatment | -0.5 year |
Severe depression | +0.9 year |
Severe anxiety | +1 year |
It goes without saying, that managing to slow down aging results in a decreased risk of life-threatening diseases - calculated from the hazard ratios
Reduction in Biological age | Risk of CVD (heart disease) | Risk of stroke | Risk of death |
1 year | -10% | -20% | -14% |
2 years | -21% | -44% | -29% |
3 years | -33% | -72% | -48% |
So, it is important to track your biological age regularly, more so with interpretable markers, so you can take actionable insights from the existing clinical research to reduce your risk of diseases before death. If you would like to dig deeper into any of this research work, check our LongevityGPT for answers with citations: https://www.askLongevityGPT.com
Comentários