Biological Age: From Philosophy to Science an Integrative Systematic Review

The objective was to analyze the scientific production on biological age and available resources to estimate it. This is an integrative systematic review carried out in PUBMED/Medline, LILACS, SciELO and Google Scholar databases, using the descriptors indexed in Health Science Descriptors (DeCS) and Medical Subject Headings (MeSH): Biological Age, Longevity and Biological Markers. Inclusion criteria: articles with full text, published in Portuguese and English, qualitative or quantitative, with definition of biological age and/or its measurement tools (biological markers). Publications that did not address the established theme, congress abstracts, annals, editorials, comments and opinions were excluded.

A total of 31 articles published between the years 1996 and 2022 were obtained. No single biomarker seems to have a safe correlation with biological age; therefore, it is fundamental to look for reliable biomarkers in the identification of individuals vulnerable to loss of functionality resulting from the age. Research includes immunological and inflammatory markers, DNA methylation, telomere length, among others, but studies involving multiple pathways interaction, practicality and low cost is the target of many of this research. Deep learning, or Deep Learning, is rapidly becoming a tool present in all areas of science and will undoubtedly be essential for accurate estimates of biological age and, consequently, for developing interventions against most chronic diseases and perhaps the very aging process.

Keywords:Biological Age; Longevity; Biological Markers

The phenomenon of world population aging has driven the search for a better quality of life for this segment, through the promotion of healthy and active aging [1]. Estimating the health of the elderly person based on their chronological age is inadequate, as senescence does not occur uniformly among people and, even in the same individual, organs and tissues have an individualized rhythm [2]. Thus, one retrospective time marker–chronological age–is insufficient to translate organic aging [3]. The possibility of reaching advanced chronological ages with good physical and cognitive performance has been widespread in different population spheres. Currently, efforts are being made to delay the aging of organisms and monitoring intrinsic capacity can be done by measuring biological age [1].

Although static chronological parameters are used by the World Health Organization (UN) and support public policies aimed at aging, Sergei Scherbov and Warren C. Sanderson [3] consider the chronological estimate outdated because it does not take into account the different intrinsic and extrinsic exposures. These authors propose biological age as representative of contemporary elderly people, who remain healthy, strong and cognitively functional. Biological age appears in an opportune and innovative way, allowing the change of the stereotype of the elderly person and contributing positively to the socioeconomic context of those who reach the 7th decade of life who may be part of the economically active population (EAP) of a country [3,4].

Biological age is defined as the indication of the body’s general state of health that establishes the degree of aging according to body functioning, associating it with genetic factors and factors extrinsic to the individual. In this sense, biological age is understood as a prospective age that proposes to analyze future repercussions on the individual’s health and, according to Sanderson and Scherbov 2010, 2013. this age corresponds to a remaining life expectancy of 15 years [3].

It is difficult to conceive the exact date on which the gaze turned beyond numerical age and the term biological age was coined. Historically, in 1996, L. Hayflick broke paradigms when he dissociated the concept of time from the aging process. In his approach, aging is not just past events, but a consequence of biological manifestations over time [5]. When analyzing the disparities between aging, it is noticed that these extrapolate the clinically pre-established chronological intervals, as in the cases of cardiovascular diseases that are presented in the literature with clinical manifestations around the 6th and 7th decades of life, it is notorious that the beginning these vascular diseases and their mortality are epidemiologically variable among individuals [6].

Understanding and mapping the physiological processes that determine the wear and tear of organisms is one of the main challenges of gerontology today, in order to adopt increasingly individualized postures in the face of clinical manifestations [5]. The important reflections on the aging of the organism in the mid- 1990s were faced with insufficient technology to measure biological age. This mismatch between advances in scientific thinking and technology lasted until 2002, a period in which defining biological age was considered by some authors entertainment and not science [7]. In the current scenario, with all the changes that have occurred at the beginning of the 21st century in the field of medicine and technology, and with the extension of longevity, the break with the chronological definition of aging gained space for discussion in the editorials of scientific journals with the discovery of biological markers of senescence [7]. Knowing the biological mechanisms that govern the aging process is important, as other physiological processes are already known and well defined, such as puberty and menopause. The current challenge has been to find the biological markers that can definitively establish the beginning of the human senescence process [8].

Parallel to childhood development milestones, it is possible to obtain aging milestones. The study of biomarkers that include changes in gene expression and concentrations of metabolites, epigenetics, telomere wear and deep learning are established in the identification of these milestones. Deep learning, one of the most sophisticated and current estimators of biological age, proposes, through samples of peripheral blood, physical activities and body shapes, to understand the health of individuals based on algorithms. This study seeks to generate repercussions in promoting the health of the elderly and contribute to advances in medical science mediated by artificial intelligence [9].

The integrative systematic review method was adhered to as it meets targets such as reviewing theories or evidence and compiling knowledge about a specific topic, promoting recognition of gaps to be filled with new research. The following steps were observed: a) formulation of the leading question; b) definition of inclusion and exclusion criteria; c) delineation of descriptors, literature search and data collection; d) critical evaluation of included studies and discussion of results; e) publication of the knowledge obtained [10,11]. The selection of articles took place from November to December 2022, guided by the question: What is biological age and how to estimate it?

Only articles that met the following inclusion criteria were selected: published in Portuguese and English, qualitative or quantitative, with definition of biological age and/or its measurement tools (biological markers), published between 1996 and 2022, available in the following databases: National Library of Medicine (MEDLINE/PubMed), Latin American and Caribbean Health Sciences Literature (LILACS), Scientific Electronic Library Online (SciELO) and Google Scholar. Publications that did not address the established theme, congress abstracts, annals, editorials, comments were excluded.

To search for articles, the following descriptors indexed to Health Science Descriptors (DeCS) and Medical Subject Headings (MeSH) were used: Biological Age, Longevity and Biological Markers. As search strategies, the descriptors were combined using Boolean operators (AND and OR). All articles selected from the keywords had their abstracts read in full, and those that were not within the established criteria and/or were repeated in the different databases. were excluded. In the identification phase, 104 articles were found. Considering the criteria of pertinence and content consistency, as well as duplicates, 72 articles were excluded. In the end, 32 scientific articles were selected [Figure 1].

After full reading of the articles that met the inclusion criteria, the information was extracted, organized and summarized. This information was organized into 1) concepts of biological age; 2) biological markers for estimating biological age.

The selected articles were summarized in the instrument for data collection. Initially, a descriptive analysis of the articles was carried out, covering the following items: title, reference, journal in which it was published, type of study and research location [Tables 1,2,3].

Table 1:Description of the articles found according to the database.

Table 2:Articles that discuss the definition of biological age.

Table 3:Articles that discuss existing biological markers.

The final sample of this review consisted of 32 scientific articles, selected by previously established inclusion criteria. Of these, eighteen (18) were from MEDLINE/PubMed, three (3) from the LILACS database, one (1) via SciELO and ten (10) via academic google. In the other databases, no qualified articles were found to fulfill the purpose of this study

The main themes displayed by the articles were: aging process and factors that promote it, “marks of aging”, biological age and biomarkers for estimating biological age. The aging process is one of the main contributors to morbidity and mortality [12]. The course of this process is preceded by two main sources: genetics and interaction with the environment [13]. Biological age is a product of these sources and indicates the state of human aging regardless of the number of years that have elapsed, as time passes equally for all living beings, but physiological changes are governed by other magnitudes. The definition of biological age has been the subject of research for decades, seeking to determine the true rate of aging, as accurate information about it will favor the discovery of intervention goals to improve health and/or delay aging [14].

Biological age definitions have become increasingly specific through scientific advances. Initially, Hayflick established differences between chronological age and biological age [5] until, in 2022, they pointed to the possibility of measuring the marks left by the biological aging process [15]. Mechanisms that contribute to aging have been summarized under the term “marks of aging” and include eleven items: loss of proteostasis, mitochondrial dysfunction, altered nutrient sensing, telomere wasting, genomic instability, cell senescence, stem cell exhaustion, alterations epigenetics and alterations of intercellular communication.

Faced with so many markers, it becomes apparently complex to identify their levels of hierarchy and their practical viability. However, although this is the ultimate goal of any scientific development, there are essential factors for understanding these biomarkers, such as realizing that different measures of biological aging may not measure the same aging processes [16] and that the term biomarker is different from a clock biological, because while the first reflects changes in the organism at the molecular or cellular level over time, the second tends to be more of a generalization of the general state of the organism [17].

The large number of studies in search of the biological clock and biomarkers of aging show the need to measure biological age in human beings [8], since age, when measured chronologically, is no longer a reliable indicator of the rates of physiological degradation of systems and organs [2]. Research proposed that biological age be considered as subject to subjectivity, but with the advancement of science, the possibility of calculating it arose, being able to determine and explain the deficits between the average life expectancy of a population and the life expectancy perceived in an individual, leaving the field of hypothesis with the discoveries of biological markers [18]. In this integrative systematic review, the main markers of aging: 1. Telomeres; 2. Epigenetic marks; 3. Biochemical compounds; 4. Deep learning.

Telomeres

The accumulation of cellular damage can produce effects on the genome. Telomeres are regions prone to degradation with advancing age [19] and the shortening of their length is understood as a strong biomarker, as it occurs in all individuals at each replication of cellular DNA. Telomere exhaustion endorses the theory: “Hayflick limit and replicative senescence” [19]. Rapid telomere shortening is linked to female aging [15], risks of cardiovascular disease [20] and Alzheimer’s disease [8]. Individuals in the age group of 60 years, with shorter telomeres, have a higher mortality potential than individuals with longer telomeres [8]. There are limitations in research related to telomeres, making their length not the main basis of a biological clock.

Epigenetic marks

Epigenetic alterations correspond to the study of structural modifications of the genome, whether by DNA methylation, chromosomal histones or other mechanisms that affect gene expression without altering the basic composition of DNA [8]. Epigenetic alterations do not constitute genetic inheritance, but regulatory mechanisms of genetics [21] and are used as biological age gauges because they correlate with the aging of the body and its relationship with various diseases, including obesity, blood glucose levels, and various causes of mortality [8].

Biochemical compounds

Small biochemical molecules, such as cholesterol, glucose, urea, calcium, are directly related to senescence [19] and can be used as biomarkers of aging. The first research based on blood biochemistry was carried out in 2016 [8] and became promising due to its ease of collection and financial accessibility, which could be used alone or complementing other analysis methods. In addition, they can be innovative, because using their results, they are able to measure the mortality risks of an individual through a clock based on artificial intelligence, where data from serological profiles of various blood tests are collected, building a database that correlate biochemical compounds and mortality risks [8].

Deep Learning

The use of artificial intelligence to profile an estimated biological age is already a reality through Deep Learning, a subfield of traditional machine intelligence, where not only data or algorithms are used, but several multiple and non-linear layers with information, as if it were an interconnected computational neural network. With its use, the estimated biological age can be useful in the process of tracing population health profiles and be an early indicator of the health status of some patients who may benefit from palliative care in the future, as a kind of monitoring that would be useful even for public policies aimed at the health network. This study is mainly based on three classes of measurements to quantify in Deep Learning algorithms: physical activities, blood samples (biomarkers) and body shapes [9].

Levine listed 5 main algorithms for estimating biological age and used the Klemera and Doubal (KD) method as the most reliable predictor to correlate with mortality. [9] Putin et al., 2016, used such markers in Deep Learning architecture, through multiple deep neural networks (DNNs) trying to assess the importance of each blood biomarker in this and they noted the five most important biomarkers that can estimate age human biological: albumin, glucose, alkaline phosphatase, urea and erythrocytes. Fischer et al [9], showed that these biomarkers are important to reveal, even in healthy people, the future risks of mortality in 5 years, with heart diseases, cancer and others, suggesting that the biomarkers can be related to aging and mortality rates.

Aging brings together the most complex combination of molecular, cellular and organic characteristics observed in organisms. With the rapid increase in older people across the world, it is a priority to develop automated ways to assess metabolic age to achieve successful aging. Biological age is a concept that a person’s actual age may differ from their chronological age. It is often seen as the true age of an individual, providing a better measure of individual life expectancy. Biological age is referred to as physiological or metabolic age, seeking to assess how different organs, physiological processes and body regulatory mechanisms are functioning and to what extent they are stable.

No single biomarker appears to have a secure correlation with biological age. It is of great interest that there are reliable biomarkers to identify individuals who are vulnerable to loss of functionality due to age. Research includes immunological and inflammatory markers, DNA methylation, telomere length, among others, but studies involving the interaction of multiple pathways, practicality and low cost is the target of many of these researches. Biological age predictors using deep neural networks became public in 2016 and gained ground in biological aging and longevity research. Deep learning or Deep Learning is a Machine Learning that employs algorithms to process data by imitating the processing done by the human brain (mathematical neurons).

Machine learning approaches are rapidly becoming a pervasive tool in all areas of science and will undoubtedly be essential for developing interventions against most chronic diseases and perhaps the aging process itself.

None.

None.