While we still lack the magic bullet for superlongevity (working on it!), it is well known that there are practical ways to extend your healthy lifespan through diet & supplements, exercise, meditation, having a healthy social network and positive outlook on life. But how are the effects of these longevity ‘interventions’ measured?
There’s no one simple answer to measuring a person’s progress toward a lengthier healthspan, but the medical field has evolved some high-quality ways to think about the problem, and integrate the various pieces of relevant information.
Biological age is a measure used in Longevity science to represent how much damage the aging process has done to your bodily systems. It is distinct from your Chronological age, or Calendar Age.
If a 40-year-old subject has a DNA methylation rate resembling that of a typical 60-year-old, this is evidence that this person’s biological age is closer to 60. This is also true for other biological factors: factors good at predicting chronological age tend to also be good at predicting biological age, because the two strongly correlate in any large enough sample. There are multiple ways to calculate biological age, from relatively simplistic (quick surveys on general health and lifestyle factors that are known to impact aging) to very complex ones based on measurements of what is happening in your DNA (methylation clocks).
In this post we briefly review some different methods of calculating biological age, and highlight Rejuve’s unique integrative approach to the challenge.
If you run a search for ‘Biological age calculator’, you will find various health & lifestyle surveys that will compute a biological age. These surveys are often based on common factors that affect health & longevity such as pre-existing conditions, smoking, diet and exercise habits, and mental health. Such surveys are often framed as a two-minute task that then leads to subscribing to a company’s website to begin to receive offers.
Others are more comprehensive, focusing on a specific body system, such as Heart Age, developed by the CDC, which measures the age of your heart in comparison with yor chronological age based on risk factors such as family history of heart disease, comorbid conditions like high blood pressure and high cholesterol, and smoking.
Mental health and psychological factors have important effects on longevity. Mental well being affects physical wellbeing in obvious ways, such as via diet and exercise behaviors, and attention to overall health condition and the motivation to seek out medical help when there are possible indicators of disease. Studies on centenarians and supercentenarians (those who have lived to 110) consistently find the theme of keeping the mind sharp and active, and continuing to have a vibrant social network. One team of researchers at Deep Longevity even came up with an aging clock based on psychological age called PsychoAge, which used deep neural networks (DNN) and anonymized surveys to estimate chronological age in combination with another calculation for self-perception of aging.
While such surveys do predict biological age and have an evidence base, surveys alone can only provide estimates, not reflect the intricacies of an individual human body.
Health/fitness trackers, often called wearables, have taken great steps to tracking and improving fitness and activity for millions of people. Wearable devices such as smartwatches, smart rings and even your mobile phone track things like steps, calories burned, heart rate variability, specific exercise routines, sleep, and even meditation. There is mounting evidence that wearables can predict biological age. These devices measure levels of physical activity with more granularity than surveys about exercise, which often overestimate or underestimate activity levels. Wearables can also help alert to early warning signs of heart problems such as atrial fibrillation and other arrhythmias. Combined with AI analysis, these devices are no longer just novelty items for fitness junkies, but actually have clinical value. In 2018, biotechnology company GERO along with researchers from the Moscow Institute of Physics and Technology (MIPT) created an AI algorithm based on physical activity data from these devices along with clinical data from a 2003–2006 US National Health and Nutrition Examination Survey (NHANES) to predict biological age, which outperformed all such models at that time. More recently, a similar study from this year again used NHANES wearables data, along with machine learning to predict biological age. Along with this biological age calculator, dubbed MoveAge, these researchers used machine learning algorithms to understand which nutrients were most highly associated with decelerated aging later in life, and more interestingly, identified an FDA-approved pharmaceutical drug associated with slowed biological aging (see linked article). There is much that is still to be learned from Wearable device data, especially when combined with other clinical data and processed by well-crafted AI algorithms.
More accurate measurements of biological age include the use of biomarkers, measurable indicators of a biological state or condition, such as the levels of specific substances in blood, urine, or soft tissues.
Much information can be gained about a person’s physiological aging by analysing their blood. Scientists can accurately guess a person’s age from the levels of certain proteins in the blood plasma. One such protein is C-reactive protein or CRP. CRP is a biomarker for inflammation in the body. While elevated CRP levels can be used to detect acute disease, it also has a well established link to aging, particularly inflammaging, a low-grade chronic inflammation associated with the aging process. Elevated CRP can also be an indicator of chronic diseases such as cardiovascular disease and chronic kidney disease, further signifying biological aging. Another marker that can be checked for in the blood is Nicotinamide adenine dinucleotide or NAD. NAD is an essential coenzyme for metabolism, and is found in all living cells. NAD levels decrease as we age, therefore an ‘older’ body has lower levels in the blood. Some longevity-focused biotech companies have even begun to create specialized blood tests specifically tailored to measure such biomarkers in regards to aging, such as BioViva.
Similarly to chronic inflammation markers, levels of certain immune system factors are predictive of biological aging. One team of scientists developed an aging clock called iAge which uses an AI algorithm developed to predict age-related inflammatory diseases and overall deterioration of the immune system related to aging. This allows for inferences to be made about health and longevity outcomes, and of course, biological age.
More accurate yet are biological age calculators based on genetic biomarkers. These include aging clocks based on factors such as telomere length. A telomere is a repeated DNA sequence region at the end of a chromosome that signifies the end of that chromosome. Telomeres are protein caps that create a barrier to protect cells from DNA damage. As our cells continue to undergo division over our lifetimes, these telomere regions become shorter and shorter, eventually exposing the chromosome and leading to cellular senescence, the ceasing of cell division. Telomere shortening and subsequent accumulation of these senescent cells is associated with the aging process, in fact, they are two of the Hallmarks of Aging, a topic that will be covered in detail in our next blog post.
One of the current most accurate methods of calculating biological age, and consequently estimating chronological age, is the Epigenetic clock or Horvath clock. Dr. Steve Horvath is an aging researcher, geneticist, and biostatistician, and bioinformatician to whom this model is credited. The epigenome is a mechanism that sits just above the genome, consisting of chemical markers that switch genes on and off. Epigenetic clocks measure DNA Methylation, or the addition of a methyl (CH3) group onto a DNA molecule. Methylation is integral to regulating gene expression. Measuring changes in DNA methylation at specific sites can more accurately discern and quantify the effects of things like smoking and obesity. The younger you are, the higher the rates of methylation, and methylation slows down proportionally with age. Horvath’s model led to spin-off models such as ‘death clock’ GrimAge, which predicts when you are likely to die. The gradual degeneration of the epigenetic modulation system, otherwise known as epigenetic alteration is yet another of the Hallmarks of Aging, to be discussed in further detail in our next blog.
The Rejuve Approach
So the big question is, how does Rejuve approach Biological Age? As you may have guessed, we apply all the above! Our upcoming app, Rejuve: Longevity, utilizes data from surveys, wearables, and medical tests. You as the data contributor decide how much information you want to provide.
The simplest input is health & lifestyle surveys. Rejuve’s surveys are based on scientific datasets such as 2017–2018 NHanes Questionaires and academic papers. Our MVP edition of Rejuve: Longevity contains over 150 data collection points, with more to be continuously added in future updates, according to best practices and research needs.
Data that is harder to obtain/more valuable to research will yield higher rewards. The more data you input, the more accurate your Biological age calculation will be, the richer insights you will get, and the more Rejuve tokens you will earn! Having a complete profile also contributes highly to research. With all of these data types, Rejuve’s multidimensional AI Framework generates a digital model of the human body using vectors. This will help unlock new discoveries that can’t be seen with just human analysis. Longevity is a new and burgeoning field of science with endless possibilities. Rejuve could possibly create a new version of the epigenetic clock, or another model entirely of accurately assessing biological age, and more importantly, halting and reversing this process at a cellular and molecular level.
We know our community is excited about wearables. As you may have read in previous blogs, the first instance of Rejuve: Longevity will pull data from Fitbit and Apple Health (including Apple watch) to bolster your insights and recommendations, and contribute to our research. We are listening to community input and are already looking into integrating other wearable devices in future iterations of the app. This is an exciting area for us, as wearable data alone combined with AI processing is already leading to discoveries.
Running through all these methods there is one common thread beyond longevity itself: artificial intelligence. Rejuve is using a unique AI suite combining deep neural embedding frameworks with symbolic reasoning using Bayesian networks and the OpenCog AGI framework. These AI tools will yield increasing insights as the Rejuve network grows and they are fed with larger and richer pools of data, but they will begin providing significant value right from the network’s launch!
Join our Telegram Community: https://t.me/rejuvecommunity
Telegram announcements: https://t.me/rejuveio
Follow us on Twitter: https://twitter.com/Rejuve_io
Like us on Facebook: https://www.facebook.com/RejuveAI/