by Ben Best
For decades biogerontologists have sought to find a biomarker of aging that would measure true biological age, and be a better predictor of life expectancy and future functionality than chronological age [EXPERIMENTAL GERONTOLOGY; 23:223 (1988)]. Such a biomarker would be a powerful tool for testing the effectiveness of interventions intended to alter lifespan, without the need to resort to large lifelong studies, which are virtually impractical for humans. Given the failure to find a biomarker of biological age after so much effort, it seems worthwhile to ask whether there are theoretical reasons for believing that no such biomarker exists. Resources devoted to a futile search could be diverted to more fruitful pursuits.
There are countless quantifiable parameters that change with age, and which could be called biomarkers of aging. The most obvious such biomarker is the graying of hair [MICRON; Van Neste,D; 35(3):1930200 (2004)]. Aging is also associated with increased serum vasopressin accompanied by decreased vasopressin effectiveness [AMERICAN JOURNAL OF PHYSIOLOGY; Tian,Y; 287(4):F797 (2004)]. Carbonyl content of protein increases substantially with age [PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES (USA); Sohol,RS; 90(15):7255 (1993)]. Lipofuscin accumulates in the lysosomes of post-mitotic cells [ANNALS NY ACAD SCI; Portas,EA; 959:57-62 (2002)]. And old memory T-lymphocytes are less able to divide when presented with an antigen than young memory T-lymphocytes [JOURNAL OF IMMUNOLOGY; Enguardea,CR; 152:3740 (1994)].
Although these and many similar biomarkers indicate changes that occur with age, none of them have been proven to correspond to biological age. None of them have been proven to better predict life expectancy and future functionality than chronological age. Why should a biological age necessarily exist or even be possible to fabricate? Why should it necessarily be possible to say that some person having a chronological age of 60 has a true biological age of 53? Wouldn't the existence of a singular biological age presume a singular underlying aging process? But there may be no singular underlying aging process. Future functionality could be very different from life expectancy [EXPERIMENTAL GERONTOLOGY; 23:223 (1988)], which would imply at least two underlying processes. If aging is the product of many forms of damage to macromolecules, tissues, and organs as some suggest [ANNALS OF THE NEW YORK ACADEMY OF SCIENCES; de Grey,A; 959:452 (2002)], there is no singular aging process or biological age.
Endogenous and exogenous damaging agents probably accelerate aging, but certainly contribute to aging-associated diseases. There is considerable overlap in the histopathology of skin photoaging and skin intrinsic aging [EXPERIMENTAL DERMATOLOGY; El-Domyati,M; 11(5):398-405 (2002)]. Victims of the tissue-specific "accelerated aging disease" xeroderma pigmentosum show photoaging in areas of skin exposed to light [MECHANISMS OF AGEING & DEVELOPMENT; Niedernhafer,LJ; 129(7-8):408-415 (2008)]. More often, however, exogenous damaging agents contributions to aging are distinguishable from intrinsic aging.
Forced expiratory volume in one second (FEV1) ranks second only to smoking as a predictor of all-cause mortality in humans [EUROPEAN RESPIRATORY JOURNAL; Young,RP; 30:616 (2007)]. F2-isoprostanes (the best available in vivo marker of lipid peroxidation) in the urine of smokers drops by more than one-third after two weeks of smoking cessation [NEW ENGLAND JOURNAL OF MEDICINE; Morrow,JD; 332:1198 (1995)], illustrating how smoking quantitatively contributes to atherosclerotic aging.
Plasma inflammatory cytokines increase with age, and are positively associated with cardiovascular mortality [IMMUNOLOGY AND ALLERGY CLINICS OF NORTH AMERICA; Bruunsguard,H: 23:15 (2003)]. In normal subjects a high-fat meal elevates plasma inflammatory cytokines significantly more than a high-carbohydrate meal [JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY; Nappo,F; 39(7):1145 (2002)]. Like smoking, a high-fat diet can contribute to blood-vessel aging. But predictability of mortality is not a sufficient condition to establish biological age.
Accelerated mortality is not accelerated aging [AGING CELL; Miller,RA; 3(2):47-51 (2004)]. Cigarette smoking and high-fat meals accelerate aging of blood vessels, whereas lead poisoning simply accelerates mortality. A biomarker of aging could distinguish accelerated aging from accelerated mortality where the forms of damage are distinct, but only if aging damage could be reduced to a single biomarker. A biomarker of aging should be applicable to nonsmokers as well as to smokers, however. Accelerated general aging — as opposed to aging in a particular tissue — could only be validated if biological age and a general biomarker of aging had been validated. It is too easy to describe accelerated mortality due to tissue damage as accelerated aging. It is too easy to impute an "accelerated aging phenotype" to the appearance of damage tissues associated with accelerated mortality.
Cellular senescence increases with age, and senescent cells contribute to cellular dysfunction and carcinogenesis of adjacent cells by secretion of growth factors and cytokines [PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES (USA); Krtolica,A; 98(21):1207212077 (2002) and MOL CELL BIOL; Campisi,J; 8(9):729-740 (2007)]. Long-term maintenance of cellular senescence requires expression of the cell cycle inhibitor p16INK4a protein. A study of rodent organs found an average 10-fold increase in p16INK4a expression with age, suggesting that this protein is a biomarker — and possible effector — of both cellular senescence and of mammalian aging [JOURNAL OF CLINICAL INVESTIGATION; Krishnamurthy,J; 114(9):1299-1307 (2004)]. Increased p16INK4a expression with age may lead to increased senescence of pancreatic β−cell stem cells in non-insulin-resistant type 2 diabetes [NATURE; Krishnamarthy,J; 443:453-457 (2006)]. — and increased stem cell senescence associated with declining neurogenesis in some areas of the brain [NATURE; Molofsky,AV; 443:448-452 (2006)]. In human T-lymphocytes, p16INK4a expression increases with age between ages 20 to 80, with twice the increase in smokers [SCIENCE; Liu,Y; 8(4):439-448 (2009)]. Removal of p16INK4a-positive senescent cells in mice delayed the onset of age-related pathologies [NATURE; Baker,DJ; 479:232-236 (2011)].
Aging doesn't just happen in blood vessels or senescent cells, however. Skin, brain, kidney, immune system, DNA, etc. all suffer damage with age — and suffer forms of damage distinguishable from that seen in atherosclerosis and senescent cells. The most vulnerable organ for an individual could determine the aging-associated disease that kills that person. A person who is relatively free of atherosclerosis and nuclear DNA-damage may die of a fatal infection due to accelerated immunosenescence.
Immunoscenescence, which increases the vulnerability to fatal infection, is significantly associated with cytomegalovirus [CURRENT OPINION IN IMMUNOLOGY; Derhovaussian,F; 21:440 (2009)]. Cancer incidence increases exponentially with age in humans up to age 80 [NATURE; De Pinho,A; 408:248 (2000)], indicative of the exponential increase in nuclear DNA damage with age. The rate of leukocyte telomere length attrition predicts cardiovascular mortality in healthy men [AGING; Epel,ES; 1(1):81-88 (2008)], and leucocyte telomere attrition rate is accelerated in obesity and insulin resistance [CIRCULATION; Gardner,JP; 111(17):2171-2177 (2005)]. Cytomegalovirus, damage to nuclear DNA, and obesity are other examples of how exogenous agents contribute to aging distinct from atherosclerosis and cellular senescence.
Aging is reputedly accelerated by genetic factors as well as by exogenous agents. But as with exogenous damaging agents, there is a multiplicity of genetic factors leading to a multiplicity of accelerated aging diseases. None of the so-called accelerated aging diseases show all of the characteristics of accelerated aging. Although Werner's Syndrome victims are characterized by bilateral cataracts, osteoporosis, severe atherosclerosis, and frequent sarcomas, there is no tendency toward Alzheimer's Disease [CELLULAR AND MOLECULAR LIFE SCIENCES; Cox,LS; 64:2620 (2007)]. By contrast, Down's Syndrome victims have a high risk of early-onset Alzheimer's Disease, along with leukemia and congestive heart failure, but there is no increase in blood pressure or most forms of cancer [HUMAN MOLECULAR GENETICS; Wiseman,FK; 18:R75 (2009)]. Hutchinson-Gilford progeria syndrome victims show early hair loss, osteoporosis, and atherosclerosis, but no cataracts or neurological problems [CELLULAR AND MOLECULAR LIFE SCIENCES; Cox,LS; 64:2620 (2007)].
Biological age could be estimated from multiple biomarkers or as a weighted-average of many forms of damage. An epigenetic study of multiple genes of 34 male identical twins between the ages of 21 to 55 was able to predict chronological age with an accuracy of 5.2 years [PLOS ONE; Bocklandt,S; 6:e14821 (2011)]. In the hippocampus of human cadavers there is a significant correlation between advanced glycation end-products and chronological age [HISTOPATHOLOGY; Sato,Y; 38:217 (2001)]. Biological age would approximate chronological age, but a biomarker of aging must be shown to better indicate biological age than chronological age. Another multi-gene study demonstrated a linear relationship between lifespan and 169 "longevity alleles" (SNPs) in the Framingham cohort [AGING; Yashin,AI; 2:612 (2010)]. Despite the linear relationship, the scattering of points about the line is very wide, suggesting poor reliability in predicting the biological age of any individual.
Validating biological age and a biomarker of aging for an individual would have to be within the context of a species or group, and would correlate with the maximum lifespan of the most long-lived 10% of that species or group. How relevant, however, are the causes of death of the other 90%? Most people die of aging-associated diseases (atherosclerosis, cancer, Alzheimer's Disease, etc.) rather than aging per se — and this is especially true of those who do not become very long-lived. Biological age might be relevant to a species or group, but not to most individuals of that species or group.
For humans, the need to validate biological age or a biomarker of aging against maximum lifespan runs against the very impracticality that motivates the search for a biomarker of aging in the first place. But how else could a biomarker of aging be validated for humans? Validating biomarkers of aging or biological age in mice would not equate with validating a biomarker of aging in humans because the mechanisms of aging appear to be different. In contrast to humans, nearly 90% of mice die of cancer [RADIATION RESEARCH; Tanaka,B; 167(4):417-437 (2007)]. Although number of cell population doublings leading to replicative senescence correlates with species lifespan [PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES (USA); Rhome,D; 78(8)L5009-5013 (1981)], and telomeres shorten more rapidly in most short-lived species [PROCEEDINGS OF THE ROYAL SOCIETY; Haussman,MF; 270:1387 (2003)], mice show no reduction of telomere length with age, due to active telomerase enzyme [SCIENCE; Mathon,NF; 291:872-875 (2001)].
People differ greatly in their vulnerability to aging-associated diseases, and the lifestyles which affect their vulnerability to aging-associated diseases. The passage of time (chronological age), the genetic condition of their organs, and exogenous damaging agents all significantly affect vulnerability to aging-associated diseases in a given individual. Yet it has been claimed that eliminating aging-associated diseases would not much alter human lifespan in the United States [SCIENCE; Oshansky,SJ; 250:634 (1990)]. This argument implies that intrinsic aging and aging-associated diseases are so closely related that not much further "squaring of the curve" is possible. It implies that there is not much intrinsic aging that acts independently of contributing to aging-associated diseases. Aging-associated diseases result in death from the most vulnerable organ or tissue. Different organs and tissues age at different rates. The widely different forms of tissue damage manifested in the various aging-associated diseases does not imply a uniform biological age associated with a single biomarker.
Calorie restriction with adequate nutrition would appear to slow aging so generally as to delay biological age, allowing for the potential of finding a single biomarker of aging. Biomarkers of calorie restriction evident in rodents have included lower temperature, lower plasma insulin, and elevated dehydroepiandrosterone sulfate (DHEAS) [SCIENCE; Roth,GS; 297:811 (2002)]. But biomarkers of caloric restriction are not biomarkers of aging. Biomarkers of caloric restriction could be imperfect biomarkers of aging because although calorie restriction reduces oxidative damage to nuclear DNA in the brain, heart, and skeletal muscle, of Fischer 344 rats, calorie restriction does not reduce oxidative damage to nuclear DNA in the liver or kidney [PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES (USA); Hamilton,ML; 98(18):10469-10474 (2001)]. As another example, rhesus monkeys subjected to caloric restriction show no decline of melatonin with age, not simply a delay in the decline of melatonin [JOURNAL OF CLINICAL ENDOCRINOLOGY & METABOLISM; Roth,GS; 86(7):3292-3295 (2001)].
In conclusion, evidence for a single biological age presumes evidence for a single underlying aging process. Aging affects so many different organs, tissues, and macromolecules in so many different ways that it seems improbable that there is a singular underlying process. Similarly, it seems improbable that a single biomarker molecule or assay could be found for a process affecting so many tissues and macromolecules in so many different ways. Moreover, even if there was a singular biological age which could be assayed by a single or composite biomarker, validating the biomarker for humans by reference to maximum lifespan could not be done in an acceptable time frame. If biological age cannot be determined for rodents, it seems unlikely that it exists for humans.
The above article was submitted to the peer-reviewed journal AGE (which is the publication of the American Aging Association) in the Fall of 2011. The journal citations were in Pub Med summary form at the bottom (as required), rather than in the linkable-in-text form that I prefer for publishing on my website. There was an abstract at the beginning, which was:
Biological age, if it existed, would provide a more precise measure of aging damage than chronological age. A biomarker of aging that would measure biological age could be useful in intervention studies. But given that aging affects so many organs, tissues, and macromolecules in so many different ways, it seems unlikely that a singular biological age exists. Accelerated mortality is not accelerated aging. Only a true biomarker of aging could distinguish between accelerated mortality and accelerated aging. Validating a biomarker of aging for humans, if one existed, would necessitate maximum lifespan studies, which are impractical. This review surveys many forms of aging and proposed biomarkers of aging to illustrate the improbability that a singular biological age exists.
In submitting my article for publication I was required to list my degrees. I don't know if this is standard practice, but it does make me wonder how much me not having a PhD or MD could serve as a filter for accepting or rejecting articles for publication. I also listed my organization as the Cryonics Institute, which makes me wonder how much concern this might raise for an organization struggling for credibility — and needing to explain the connection to hostile critics of cryonics inside and outside the American Aging Association.
I have attended many annual meetings of the American Aging Association because of my interest in biogerontology (life extension). I created, and have been the main person maintaining, the American Aging Association entry on Wikipedia. I have met Don Ingram (the Editor-in-Chief of AGE) many times at American Aging Association meetings, but I don't know how well he associates my name with my appearance. Dr. Engram sent me the following rejection letter concerning the article above which I had submitted for publication:
Dear Mr. Best:I have received and read your manuscript (JAAS594), "Biological Age: A Quest without a Quarry?" that you submitted to AGE. While the paper clearly relates to an important topic, I have decided that you have not provided sufficient coverage of the subject. Primarily you have not cited many other important papers that have addressed this issue. Consideration of other approaches is crticial to evaluating your proposal.
I would like to thank you very much for forwarding your manuscript to us for consideration and wish you every success in finding an alternative place of publication.
When there are deficiencies in a submitted paper it is common practice for journal editors to request that these deficiencies be corrected, and the paper be resubmitted. The impression I get from Dr. Ingram's message is "Go away and stay away". I have not submitted the article to another journal for publication, and have no plans to do so. Possibly this leaves me yet another embittered, resentful author who has written a lousy article for publication that was rejected. I do not have the objectivity to claim that I wrote an excellent article that was rejected for petty reasons by the Editor-in-Chief, who did not even send the article out for peer-review. I do know that Don Ingram has considerable expertise in the subject of biomarkers of aging, and I believe that this expertise may be associated with many biases. I also know that my article is written to challenge the assumptions of those who have spent their careers searching for biomarkers of aging. Dr. Ingram gave as a reason for rejection the fact that I did not include notable references to the existing literature on biomarkers of aging. I included citations to a few key papers, but for the most part I do not think the literature on biomarkers of aging has much additional to say — or even much additional to refute — that I did not already cover in my article.
In particular, I did not reference the 2001 paper on biomarkers of aging by Richard Miller because I did not think that doing so would add anything to my article. I acknowledge that this could have been a political mistake on my part. Richard Miller is an active member of the American Aging Association and he regularly attends the annual meetings of that organization. In fact, I had briefly argued with Miller about biomarkers of aging at the 2011 annual meeting. I find him to be offensively condescending and arrogant. Before I knew Dr. Miller very well I visited him at the University of Michigan to probe his biogerontology knowledge and to ask him for a job. I am sure that he thought that my biogerontology questioning was just a ploy to get a job, but he does not appreciate the intensity of my interest in biogerontology nor the fact that my wanting to work for him was associated with that interest. Given Miller's status in the field of aging biomarkers, I expect that Ingram would have felt obliged to send my article to Miller for review, noting the fact that I had not referenced Miller's 2001 paper (although I did reference Miller's paper on accelerated aging in my review). I expect that Miller would have nixed my article, I expect that Ingram had the same expectation, and I expect that Ingram deemed it was better to reject my article than have to explain to Miller why he had not done so.
Of course, this is all speculation on my part. I may simply have written a bad article. Again, it is hard for me to be objective about this matter. I am frankly not convinced that the thesis of my article is correct. I can see both sides of the issue, although I believe that I have presented a side of the story that had not been raised before, and which merits discussion. I regret not having cited and discusssed the relevance of Miller's 2001 paper, and not simply for political reasons — in sympathy with the comments in Ingram's rejection letter. As penance, I created a Wikipedia entry on biomarkers of aging which included reference to Miller's 2001 paper among the few papers I cited.