It’s estimated by 2050, there will be 2.5 times more people aged 65 years and above, than those aged 4 years old or less.(1) The ramifications of this demographic change will be widespread and significant. With increased pressure on global healthcare systems to treat age-related conditions (e.g. cardiovascular, neurodegenerative and respiratory diseases), governments will need to effectively utilize technological solutions to cope with the rising demand. For example, the life expectancy of women in developed countries has increased dramatically, from an average of 45 years old in 1840, to 85 in 2015.(2) With such a leap in human lifespan occurring over a period of a mere 200 years, it’s clear that longevity is significantly influenced by environmental factors.
There’s a great deal of variability observed in both the health state and lifespan of individuals. At one end of the spectrum of elderly we have healthy centenarians (people that are 100 years old and above), and at the other, 60 years olds that suffer from multiple age-related diseases. Researchers have tried to establish the genetic causes for this variability, using animal models and genomics techniques to locate genes with the potential for extending lifespan.
Early theories on ageing proposed that age-related changes were the result of somatic DNA mutations, leading to an accumulation of structural errors in proteins and a loss of cellular functions. Studies later revealed that the rate of mutations in fact differ between both tissue types and organs, posing challenges to research using data obtained from bulk tissues.(3) Contradictory findings in studies aiming to link somatic mutations and lifespan have further brought into question whether mutation rates are an accurate assessment of a person’s overall “biological age.”
Due to the heterogeneity of mutation rates, a number of differing approaches have been employed to try and estimate overall biological age. Research has focused on identifying lifespan regulating gene loci, or loci that could be involved in age-related traits and disease. For example, hypothesis-free approaches applied to linkage studies of longevous families (e.g. GEHA study) and long-living sibling pairs have uncovered insights into healthy ageing. Recently, Genome wide association studies (GWAS) have been used to map age-related disease susceptibility and in animal models, proteome maintenance has been found to correlate with longevity.(4)
Organisms undergo phenotypic modifications with ageing, for example reductions in muscle tone, visual acuity and flexibility. Whilst several gene candidates have been linked to ageing, researchers have found it difficult to establish connections between genes and phenotypic starting points for inquiry. For example, how can we investigate ageing when it’s not entirely clear exactly which physiological changesrepresent the early stages of an age-related health decline? What we call “ageing” would perhaps be better thought of as a broad and diverse collection of molecular, cellular and anatomical changes that occur with advancing age.
Many genetic disorders, such as early-onset osteoarthritis, express phenotypes that could be considered to be “ageing-like” but we’ve come to recognize that most Mendelian Disorders are in fact not related to the gradual degenerative changes observed in ageing populations. Ageing then, should be considered to be a multi-factorial functional decline that increases the mortality risk of individuals; possibly as the result of an accumulation of molecular damage from spontaneous biochemical errors and free radicals.(5)
Whilst genetics undoubtedly has an impact on ageing, it is likely that ageing-related processes are more heavily influenced by a complex interplay of phenotypes along a patient’s life course. These include early-life factors such as birth weight, mid-life factors such as lack of exercise, and late-life complex health traits such as Cardiovascular Diseases (CVD). In twin studies, for example, it’s estimated that genetics accounts for only 20-30% of the lifespan variation observed.(6)
One clue into the far-reaching molecular impacts of environmental factors was observed in studies investigating the Dutch Hunger Winter. In this event, early adverse prenatal exposure to famine resulted in persisting epigenetic effects observed in mid-life (i.e. a reduction in DNA methylation of IGF2).(7) Epigenetic changes can be measured by the amount of DNA methylation occurring. For example, if DNA methylation at genomic loci match chronological ageing then these changes can be used to construct a prediction of a person’s “biological age.” DNA methylation age (DNAmAge) has been demonstrated to act as an effective age predictor.
Alongside genomics, transcriptomics (the study of mRNA production) can be used to identify ageing-related regulation of pathways. Genes relating to inflammation are up-regulated in ageing, which is consistent with other findings that have suggested that ageing is characterized by chronic low-grade inflammation.
Age-related regulation of gene expression are commonly considered to be detrimental, but some researchers have suggested that these changes could in fact be compensatory to other underlying processes. Disentangling the different cause-consequence relationships in gene expression regulation can be a tricky proposition. Likewise, establishing a “transcriptome age” of an individual is highly challenging as the amplitude of mRNA regulation is typically of a small volume, leading to a lack of reproducibility across studies.
In order to understand the molecular changes underpinning ageing, researchers have turned to identifying and interpreting molecular biomarkers (e.g. blood-based metabolites), a range of disease-specific molecular indicators that can be measured by using metabolic profiling techniques.
In a study comparing middle-aged offspring of long-lived subjects, poor metabolic health was found to inversely associate with familial longevity.(8) Using high-throughput metabolic profiling techniques, public health researchers now have the opportunity to screen population-size cohorts, and biobanks, for biomarkers linked to age-related diseases and all-cause mortality.
As there is so much variability in ageing between individuals, some physicians have suggested that precision medicine could provide healthcare systems with a “biological age scoring system.” This would require the development of a diagnostic technique that measures biomarkers that indicate the pace of ageing, and could be used to provide an overall estimate of a person’s metabolic age. Clinically, this hypothetical system would allow physicians to monitor, and potentially slow, the pace of an individual’s ageing. Interventions such as lifestyle changes (e.g. increased exercise) or medications could be prescribed, with any resulting changes in a patient’s biological age monitored in real time.
There have been a number of candidates for biomarkers that can indicate biological age, for example single-marker indicators, such as the much-publicized leukocyte telomere length (LTL). Whilst telomere length has been linked to ageing in several studies, it is more likely that a multi-biomarker algorithm will form the base of any effective biological age score.
Metabolic biomarkers offer us strong molecular predictors of age-related health and mortality, and can potentially be combined with other predictors. All-cause mortality is particularly useful in ageing research as it has the potential to facilitate better risk-prediction and treatment of high-risk patients. For example, four circulating biomarkers (alpha-1-acid glycoprotein, albumin, very-low lipoprotein particle size and citrate) were found to predict the risk of all-cause mortality over a 5-year follow up.(9) Other techniques such as urine-based metabolomics platform have also been used to generate a biological age predictor.(10)
There is evidence of other clinically important biomarkers for metabolic health, or indicators of “biological youthfulness.” In a 13-week lifestyle intervention study in older adults (the Growing Old Together study), it was found that a 25% reduction in energy balance resulted in a number of metabolic changes that were similar to previously described associations with low risk of type 2 diabetes and cardiovascular disease. These changes included a decrease in metabolites involved in inflammation, apolipoproteins, branched-chain amino acid leucine and aromatic amino acid tyrosine. Interestingly, differences were observed in HDL-related metabolites between male and female participants, suggesting that the metabolic effects of ageing are also influenced by gender.(11) For example, metabolic profiling of menopausal women revealed alterations in lipids and amino acid concentrations that correspond with increased CVD risk.(12)
In studies examining long living family sibling pairs and centenarians, a number of metabolic differences were observed when compared to controls. These included: low glucose levels, maintenance of insulin sensitivity, low free triglycerides, high adiponectin and large LDL/HDL particle size.(13)
With high-throughput metabolomics screening (such as Nightingale’s platform) and a multi-biomarker biological ageing score, there is a possibility to provide insights into the impact biological ageing has on the risk of developing chronic diseases. Measuring biomarkers over different time scales allows us to link molecular vulnerability with clinical outcomes, facilitating effective risk scoring and patient group stratification based on biological, not chronological, age.(14,15)
The benefits of establishing a biological age for patients are significant. Medications are typically developed and tested in younger patients, due to the heterogeneity of health state and response in elderly populations. By developing a better understanding of ageing we can establish the effect of interventions and treatments. This allows for the development of medications that can be prescribed to target groups based on biological age and with improved efficacy.
It is clear that we are closer than ever to a change in how we medically perceive age. Perhaps instead of celebrating the chronological passing of another year, we’ll be cheering on the “youthfulness” of our metabolites!
1. João Pinto da Costa et al. Analytical tools to assess aging in humans: The rise of geri-omics. TrAC Trends in Analytical Chemistry. (2016) 80:202-212 https://www.sciencedirect.com/science/article/pii/S0165993615300182
2. Eline Slagboom et al. Phenome and genome based studies into human ageing and longevity: An overview. Biochimica et Biophysica Acta. (2017) pii: S0925-4439(17)30332-0. https://www.ncbi.nlm.nih.gov/pubmed/28951210
3. Alessandro Cellerino et al. What have we learned on aging from omics studies? Seminars in Cell and Developmental Biology. (2017) 70:177-18 https://www.ncbi.nlm.nih.gov/pubmed/28630026
4. S.B. Treaster et al. Superior proteome stability in the longest lived animal. Age (Dordr). (2014) 36(3):9597 https://www.ncbi.nlm.nih.gov/pubmed/24254744
5. Matteo Tosato et al. The aging process and potential interventions to extend life expectancy. Clinical Interventions in Aging. (2007) 2(3):401-412 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2685272/
6. T.J.C. Polderman et al. Meta-analysis of the heritability of human traits based on fifty years of twin studies. Nat Genetics (2015) 47(7):702-709 https://www.ncbi.nlm.nih.gov/pubmed/25985137
7. Krista Fischer et al. Biomarker Profiling by Nuclear Magnetic Resonance Spectroscopy for the Prediction of All-Cause Mortality: An Observational Study of 17,345 Persons. PLoS Med. (2014) 11(2): e1001606. http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001606
8. J. Hertel et al. Measuring Biological Age via Metabonomics: The Metabolic Age Score. Journal of Proteome Research (2016). 5;15(2):400-410 https://www.ncbi.nlm.nih.gov/pubmed/26652958
9. Bastiaan T. Heijmans et al. Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proceedings of the National Academy of Sciences of the United States of America. (2008) 105(44):17046-17049 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2579375/
10. A. A. Vaarhorst et al. Lipid metabolism in long-lived families: the Leiden Longevity Study. Age (Dordr). (2011) 33(2):219-227 https://www.ncbi.nlm.nih.gov/pubmed/20811950
11. O. van de Rest et al. Metabolic effects of a 13-weeks lifestyle intervention in older adults: The Growing Old Together Study. Aging (Albany NY). (2016) 8(1):111-126 https://www.ncbi.nlm.nih.gov/pubmed/26824634
12. Kirsi Auro et al. A metabolic view on menopause and ageing. Nature Communications. (2014). doi:10.1038/ncomms5708 https://www.nature.com/articles/ncomms5708
13. C. A. Wijsman et al. Familial longevity is marked by enhanced insulin sensitivity. Aging Cell. (2011) 10(1):114-121 https://www.ncbi.nlm.nih.gov/pubmed/21070591
14. Ville-Petteri Mäkinen and Mika Ala-Korpela. Metabolomics of aging requires large-scale longitudinal studies with replication. Proceedings of the National Academy of Sciences of the United States of America. (2016). 113(25):E3470 http://www.pnas.org/content/113/25/E3470
15. Marian Beekman et al. Classification for Longevity Potential: The Use of Novel Biomarkers. Frontiers in Public Health. (2016). 4:233 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5083840/