Jump to content
Powered by

Looking for a formula to determine biological age

The Europe-wide "MARK-AGE" project involving scientists from 14 countries commenced on the 1st April 2008. Scientists have joined forces to investigate the factors involved in ageing. They are using standardised questionnaires and analysing data acquired from physical and biochemical examinations of the blood and urine from 3,700 volunteers. The researchers hope to be able to find better ways of preventing diseases such as Alzheimer’s, osteoporosis or arthrosis, all of which tend to occur mainly in the latter part of a person’s life. The team of experts includes the medical expert Professor Alexander Bürkle and the mathematician Prof. Michael Junk who is responsible for the biostatistical analysis of the data acquired. The two researchers from the University of Constance talked to Michael Statnik about their search for a formula to determine the biological age of humans.

Prof. Bürkle, what is the current state of the project?

Bürkle: We are still in the set-up phase of the five-year project. We are currently establishing biobanks for storing the samples, a central database and the standard protocols. We are also organising the shipment of samples, which is a very complex procedure, and are currently working on the first “dry run”.

Can you highlight the factors that induce and enhance the ageing processes, from a scientific perspective?

Prof. Dr. Alexander Bürkle coordinates the MARK-AGE project. © A. Bürkle

Bürkle: Ageing depends on the genes that people inherit from their parents. This genetic makeup cannot be altered. But it also depends on lifestyle, which fortunately, can be changed. Factors such as tobacco use, excessive alcohol consumption, poor diet, being overweight, lack of exercise and sleep and many other factors affect the ageing process. We want to focus on the analysis of different parameters in our cross-sectional study. This includes simple basic examinations such as body size, weight, blood pressure, creatinine and cholesterol values, lung capacity, muscle power, grasping strength, memory tests, blood samples and so on.

The project involves 26 international research groups. Is there a wide variety of opposing theories?

Bürkle: No, not at all. Our project is not based on one of many possible ageing theories, but is purely of an empirical nature. But it is possible that the results of our project will later affect the development of theories. Our path is clear: We hope that MARK-AGE will lead to a combination of the most promising biomarkers, which will then allow further gerontological studies to be carried out.

What might a formula for the biological age of humans look like?

Bürkle: How about: biological age = 0.2 x CRP [in mg/dl] + 0.04 x log10 of abdominal girth [in cm] + 0.5 x the second power of telomer length [in kb] - 0.7 x the square root of the DNA repair capacity [in %]… And this could be continued with the many individual parameters that affect the biological age of humans.

Junk: Such linear combinations of transformed biomarkers have been proposed in the literature. But it is worth noting that the term “biological age” is still very vague. I see the mathematical specification as another major aim of the MARK-AGE project.

Prof. Junk, what are the criteria for promising biomarkers?

Junk: An important criterion is certainly the correlation with a person’s chronological age, but there are more criteria than this. In the MARK-AGE project, we will collect data from a group of very long-lived people and their normally ageing life partners. We will also collect reduced amounts of data on people that age faster than normal. An important biomarker criterion will be that the biomarkers have to be able to clearly distinguish the differently ageing groups from one another.

Can you give me a few examples of parameters that affect the biological age of humans and that are being analysed as part of the new project?

Bürkle: We will analyse “classical” biomarkers, for example C-reactive protein, hormone-based and DNA markers in order to see how significant they actually are. C-reactive protein is a known plasma protein that is found in higher quantities in inflammatory reactions. The hypothesis is that the ageing of organisms is associated with a general, low-grade inflammatory process, which might potentially be the reason why we get older. We can easily test this on our volunteers using C-reactive protein. Such an analysis has never before been carried out on such a large proband collective.

Will you also use completely new measurement parameters and methods?

Bürkle: Yes. Among the 150 individual parameters used there are a number of new factors that would appear to have an impact on ageing. For example, the length of telomeres, DNA repair capacity, expression and activity of the DNA repair enzyme poly(ADP-ribose)-polymerase-1 (PARP-1), DNA methylation patterns of six selected genes, the accumulation of mutations in mitochondrial DNA, alterations of N-glycosylated proteins in the plasma, the plasma levels of the ApoJ protein, isoprostanes (as the markers of oxidative stress), cellular glutathione (a cellular antioxidant) and many more. A number of new measurement methods will be developed in the course of the project.

What new findings is the investigation of telomere length likely to discover?

Bürkle: Telomeres are the end pieces of chromosomes consisting of a 6 nucleotide sequence (TTAGGG) that is repeated more than a thousand times. We will use fluorescence-labelled probes that are complementary to these telomere sequences and can form perfect base pairs with TTAGGG (i.e. AATCCC, which is however antisense, i.e. it must be written as CCCTAA). These specific probes can bind to all TTAGGG sequences available if the cells and chromosomes are specifically prepared. The signal acquired with a fluorescence microscope is proportional to the number of TTAGGG stretches and hence a measure of telomere length. A lot of data from animal experiments show that the enhanced loss of telomere segments at the end of the chromosomes and the weakening of specific DNA repair processes considerably accelerates the ageing of the chromosomes. In addition, it is known that human telomere length decreases with increasing age. It is suspected that the DNA repair function also decreases with age.

Why do you use the DNA repair enzyme poly(ADP-ribose)-polymerase-1 (PARP-1) in your biochemical and molecular investigations? What does the increase of the enzyme's activity tell you?

Prof. Dr. Michael Junk will analyse the data. © M. Junk
Bürkle: This enzyme is found in the cell nucleus and produces poly(ADP-ribose) which is required for the efficient repair of damaged DNA, in particular for “base excision repair”, which is one of six major repair pathways in human cells. The cells which we will investigate – in this case a subgroup of white blood cells – will be fixed and the cell membrane made permeable. The cells are fed with all they need to produce as much poly(ADP-ribose) as possible, i.e. NAD+ as substrate, short DNA segments as surrogate for broken DNA strands, and a suitable buffer. The poly(ADP-ribose) produced in vitro will then be bound with a specific antibody. This antibody is detected with a fluorescence-labelled secondary antibody and the strength of fluorescence quantitatively determined with a flow-through cytometer. The greater the fluorescence intensity, the greater the activity of PARP-1.

Does your survey include parameters that differentiate 35-year-olds from 74-year-olds?

Bürkle: In the empirical study, people over 60 have to answer some questions which are not relevant to younger people. These questions include for example: do you need help to undress, eating or when going to the toilet?

Is family background important for your analyses?

Bürkle: Due to different genetic makeup and different lifestyles, we expect considerable differences between the different European countries where the volunteers are being recruited. One exclusion criterion is that all volunteers need to have spent at least half of their lives in the country where they live at the time of the investigation.

How general can tests involving 3,700 volunteers be?

Bürkle: The number of volunteers is very large and represents the situation in Europe. However, it is clear that our results do not allow conclusions to be drawn on populations in Africa, Asia or Latin America. Additional studies are necessary to make statements on the biological ageing of these populations. Our European study might have a pilot role because we will be able to identify unsuitable and redundant tests, which will not then be used in future tests. This considerably simplifies future investigations at the same time as reducing costs.

Junk: One has to take into consideration that the cross-sectional data reflect systematic effects that, for instance, are due to the fact that older volunteers had a different diet when they were young, or that the pollution burden has systematically changed due to technical developments or changing political conditions. Therefore, the median value of certain biomarkers in the class of today’s 70-year-olds will most likely not tell us much about the median value of the same markers of today’s 40-year-olds in about 30 years’ time. That is why the MARK-AGE project includes plans to carry out a longitudinal study. The acquisition of data from around 12 per cent of all the project participants in about four years’ time will provide additional information about the temporal behaviour of the biomarkers under investigation. It is also assumed that the biological age is, on average, steadily increasing with the chronological age of people.

Do you have plans to use the new findings and turn them into products, either now or in the future? What is the purpose of including companies in the project?

Bürkle: Yes, we have plans to turn our findings into practical applications, for example in preventive medicine, where (still) healthy people with accelerated biological age will be screened. These people will then be medically examined and the preventive treatment of age-associated diseases be put in place.

Professor Junk, which mathematical and statistical methods are you using in the project?

Junk: Initially, we will use test methods that have previously been used for similar studies. Major keywords are multivariate regression, single value partition or principal component analysis, as well as decision and regression trees. In addition, we will also use modern approaches from the fields of machine learning and data mining. A concrete objective is the reduction of the number of biomarkers required to determine the biological age to a minimum at the same time as giving as much information as possible. We will have to identify markers that correlate well with the age of the volunteers, and identify those that are only minimally coupled with each other and provide non-redundant information about the state of an organism.

Do you have an idea about the significance of parameters for determining a person’s biological age?

Junk: This is difficult to say. But we can assume that some markers will have a major significance in important body systems such as metabolism, cardiovascular system or immune system. An exciting question is also whether the DNA- or protein-based markers will provide additional information. But this question has to be answered using systematic correlation analysis. 

Website address: https://www.gesundheitsindustrie-bw.de/en/article/news/looking-for-a-formula-to-determine-biological-age