Validation of biomarkers of aging

In the paper

“Regardless of the development strategy, most current biomarkers of aging predict aging-related outcomes and identify factors associated with the pace of aging, in retrospective epidemiological studies 1-7.”

What the sources say

• Partly supported

  The Clinical applications section says many studies use aging biomarkers to predict healthspan, lifespan, and age-related conditions. But Ongoing challenges and directions for future research distinguishes biological-age biomarkers from measures of the current pace of aging; the cited passages do not establish that most biomarkers identify factors associated with that pace in retrospective epidemiological studies. Page numbers are unavailable in our copy. (LLM, conf=0.84)

  cites [1] Moqri, M. et al. Biomarkers of aging for the identification and evaluation of longevity interventions. Cell 186, 3758–3775 (2023). sourceExperimental — accuracy not yet measured
• ?Could not verify

  The Introduction (page unavailable) says that methylation markers correlate with age and are used to build epigenetic clocks. The Discussion (page unavailable) describes earlier biomarkers as predictive models, but neither passage says that most biomarkers predict aging-related outcomes or identify factors associated with the pace of aging in retrospective epidemiological studies. Searches for those outcomes and that study setting found no supporting passage in our copy. (LLM, conf=0.72)

  cites [2] Moqri, M. et al. PRC2 clock: a universal epigenetic biomarker of aging and rejuvenation. Preprint at bioRxiv https://doi.org/10.1101/2022.06.03.494609 (2022). sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction (page unavailable) says second-generation DNA-methylation clocks estimate morbidity or mortality risk and the rate of biological aging; it also says clocks respond to interventions. It does not establish that most current aging biomarkers both predict aging-related outcomes and identify associated factors specifically in retrospective epidemiological studies. The Introduction instead describes first-generation clocks as designed to predict chronological age. (LLM, conf=0.84)

  cites [3] Poganik, J. R. et al. Biological age is increased by stress and restored upon recovery. Cell Metab. 35, 807–820 (2023). sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction (page not provided) says several DNA-methylation clocks estimate mortality risk, while the Discussion reports that GrimAge2 is associated with age-related outcomes. The Results section on diet and clinical biomarkers describes cross-sectional associations between GrimAge measures and dietary, clinical, and lifestyle variables. These passages support the claim for GrimAge and some related clocks, but do not establish that most current aging biomarkers both predict outcomes and identify factors associated with the pace of aging. (LLM, conf=0.85)

  cites [4] Lu, A. T. et al. DNA methylation GrimAge version 2. Aging 14, 9484–9549 (2022). sourceExperimental — accuracy not yet measured
• Supported

  The Results (Proteomic signatures of age; page not provided) report that the proteomic age measure predicts faster accumulation of chronic diseases. The Discussion (page not provided) reports that age-associated proteins predict multimorbidity and mortality over follow-up and may help identify people aging faster than their chronological age. This paper supports the citing sentence as an example of an aging biomarker studied in an epidemiological cohort; it does not need to establish the claim about ‘most’ biomarkers on its own. (LLM, conf=0.86)

  cites [5] Tanaka, T. et al. Plasma proteomic biomarker signature of age predicts health and life span. eLife 9, e61073 (2020). source
• Partly supported

  The Results passage (Additional information; page not provided) reports that a metabolite score was associated with longevity in women, supporting the claim that a biomarker can predict an aging-related outcome. The paper does not establish what most current biomarkers do or identify factors associated with the pace of aging. Its Statistical analysis section describes outcomes tracked from baseline until death or last contact, rather than a retrospective study. (LLM, conf=0.86)

  cites [6] Balasubramanian, R. et al. Metabolomic profiles associated with all-cause mortality in the Women’s Health Initiative. Int. J. Epidemiol. 49, 289–300 (2020). sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section (page unavailable) says DNA-methylation clocks predict age-related conditions, including lifespan. It does not say that most current aging biomarkers identify factors associated with the pace of aging, or that they do so in retrospective epidemiological studies. (LLM, conf=0.85)

  cites [7] Sarkar, T. J. et al. Transient non-integrative expression of nuclear reprogramming factors promotes multifaceted amelioration of aging in human cells. Nat. Commun. 11, 1545 (2020). sourceExperimental — accuracy not yet measured

In the paper

“Currently, epigenetic markers are the most commonly proposed and investigated type of composite biomarkers 1.”

What the sources say

• Partly supported

  The Clinical applications section says epigenetic clocks are increasingly used as candidate aging biomarkers in longevity-related trials. The Classifications section describes molecular biomarkers as the most developed class and includes epigenomic measures, but it does not say epigenetic markers are the most commonly proposed and investigated type of composite biomarker. Page numbers are not provided in our copy. (LLM, conf=0.82)

  cites [1] Moqri, M. et al. Biomarkers of aging for the identification and evaluation of longevity interventions. Cell 186, 3758–3775 (2023). sourceExperimental — accuracy not yet measured

In the paper

“For example, the ApoE4 allele is the strongest risk factor for Alzheimer’s disease risk in white populations, but the association is substantially weaker in African American and Hispanic populations 67.”

What the sources say

• ?Could not verify

  The passages in “Supporting discovery through good data management” discuss gene-expression data and how to find and reuse it, not Alzheimer’s disease risk. Searches for ApoE4, Alzheimer’s disease, and population-specific risk returned no relevant passage in our copy. (LLM, conf=0.78)

  cites [67] Wilkinson, M. et al. The FAIR Guiding Principles for scientific data management and stewardship. Sci. Data 3, 160018 (2016). sourceExperimental — accuracy not yet measured

In the paper

“For example, Huan et. al 32,33 and Deelen et al 34 reported increased mortality risk (HRs of 1.85 and 2.73) for their epigenetic and metabolomic biomarkers, respectively.”

What the sources say

• ?Could not verify

  Section 13, “Measuring the Value of the Biobank,” discusses studies using Biobank data but does not report mortality risks for epigenetic or metabolomic biomarkers. Searches for Huan, Deelen, and the hazard ratios 1.85 and 2.73 found no passage addressing the claim in our copy. (LLM, conf=0.76)

  cites [32] Boutin, N. T. et al. The evolution of a large biobank at Mass General Brigham. J. Pers. Med. 12, 1323 (2022). sourceExperimental — accuracy not yet measured
• Partly supported

  The Results (Mortality prediction model; page not available in our copy) report increased mortality risk for a CpG-based epigenetic score, but not the attributed HR of 1.85. They give an all-cause mortality HR of 2.21 in ARIC, or 1.68 after adjustment for age, sex, and clinical risk factors. This verdict addresses only the Huan et al. attribution, not Deelen et al. (LLM, conf=0.88)

  cites [33] Huan, T. et al. Integrative analysis of clinical and epigenetic biomarkers of mortality. Aging Cell 21, e13608 (2022). sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction (page unavailable) reports that metabolic biomarkers in a study cited to Deelen were associated with all-cause mortality risk. It does not report an HR of 2.73, an HR of 1.85, or a finding by Huan; the References identify the methylation study it discusses as Zhang et al. (LLM, conf=0.83)

  cites [34] Eiriksdottir, T. et al. Predicting the probability of death using proteomics. Commun. Biol. 4, 758 (2021). sourceExperimental — accuracy not yet measured

In the paper

“So far, relatively few studies have compared individual (composite) biomarkers across multiple cohorts, or multiple biomarkers across the same cohort using standardized and equivalent measurement units that make them fully comparable 32,35,36.”

What the sources say

• ?Could not verify

  Section 13 discusses the MGB Biobank’s value and examples of research using its data, but not how often studies compare composite biomarkers across cohorts or within a cohort. Searches for composite biomarkers and standardized, equivalent measurement units found no passage addressing that claim. (LLM, conf=0.76)

  cites [32] Boutin, N. T. et al. The evolution of a large biobank at Mass General Brigham. J. Pers. Med. 12, 1323 (2022). sourceExperimental — accuracy not yet measured
• Partly supported

  The Results describe comparisons of biomarker-based mortality prediction across cohorts, so this paper is an example of the kind of study cited. But the Methods say biomarker values were scaled separately for each study, and the Discussion says the resulting score needs concentration units for clinical use. The paper does not establish that relatively few studies exist or that its measurements are fully comparable in equivalent units. (LLM, conf=0.85)

  cites [35] Deelen, J. et al. A metabolic profile of all-cause mortality risk identified in an observational study of 44,168 individuals. Nat. Commun. 10, 3346 (2019). sourceExperimental — accuracy not yet measured
• Partly supported

  The Discussion (page unavailable) says that, to the authors’ knowledge, metabolomic aging biomarkers had not been evaluated in other studies, supporting a gap for that specific class. The Methods > Statistical Analysis (page unavailable) describes standardizing biomarkers to improve comparability, and the Discussion describes comparisons across two cohorts. These passages do not establish the broader claim that relatively few studies have made either kind of comparison using equivalent measurement units. (LLM, conf=0.82)

  cites [36] Kuiper, L. M. et al. Epigenetic and metabolomic biomarkers for biological age: a comparative analysis of mortality and frailty risk. J. Gerontol. A Biol. Sci. Med. Sci. 78, 1753–1762 (2023). sourceExperimental — accuracy not yet measured

In the paper

“For example, many existing models assume a linear relation between biomarkers of aging and the likelihood of aging outcomes throughout the lifespan, while recent studies have discovered multiple examples of non-linearity5,12,49.”

What the sources say

• Partly supported

  The Results (‘Differential protein expression across age-span’) and Discussion report non-linear changes in protein expression with age, including a peak at 78 years, and describe linear models used to identify age-associated proteins. They do not establish multiple non-linear relationships between biomarker levels and the likelihood of aging outcomes throughout the lifespan; the non-linearity reported concerns protein expression across ages. Page numbers are unavailable in our copy. (LLM, conf=0.84)

  cites [5] Tanaka, T. et al. Plasma proteomic biomarker signature of age predicts health and life span. eLife 9, e61073 (2020). sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section (‘Quantification of proteomic changes across the lifespan’; page unavailable) reports nonlinear waves of plasma-protein changes with age, and the Discussion says linear modeling misses some of those changes. The paper does not establish that the relationship between aging biomarkers and the likelihood of aging outcomes is nonlinear; its Methods instead describe linear models for associations with clinical and functional variables. (LLM, conf=0.86)

  cites [12] Lehallier, B. et al. Undulating changes in human plasma proteome profiles across the lifespan. Nat. Med. 25, 1843–1850 (2019). sourceExperimental — accuracy not yet measured
• ?Could not verify

  The ‘Challenges in defining healthspan’ section describes health as changing dynamically over life, but it does not discuss linear assumptions in biomarker-based models or report studies finding nonlinear relationships with aging outcomes. Searches for the claimed linear relation and for nonlinear biomarker–outcome models found no passage addressing either point; page numbers were unavailable in our copy. (LLM, conf=0.76)

  cites [49] Kaeberlein, M. How healthy is the healthspan concept? GeroScience 40, 361–364 (2018). sourceExperimental — accuracy not yet measured

In the paper

“Similarly, metabolomic or proteomic data could be collected from plasma or serum, leading to different data distributions 34; measured using different technologies (e.g. mass spectrometry or aptamer-based assays); or tagged using different nomenclature39.”

What the sources say

• ?Could not verify

  Citation 39 is attached to the claim that data may be tagged using different nomenclature. The Methods (Protein measurements; page unavailable) identify proteins by UniProt ID and describe aptamer-to-protein mappings, but do not say that datasets use different nomenclatures. A search for nomenclature and protein-labeling differences found no passage addressing that point. (LLM, conf=0.72)

  cites [34] Eiriksdottir, T. et al. Predicting the probability of death using proteomics. Commun. Biol. 4, 758 (2021). sourceExperimental — accuracy not yet measured
• ?Could not verify

  The Introduction and Methods examine DNA methylation in cord and whole blood measured with 450K and EPIC BeadChips. A search of our copy for plasma, serum, metabolomic and proteomic data found no passage addressing whether plasma- versus serum-derived data have different distributions. The citation is attached to that specific point, not to the sentence’s separate points about technologies or nomenclature. (LLM, conf=0.76)

  cites [39] Olivia, S. et al. Comparison of DNA methylation measured by Illumina 450K and EPIC BeadChips in blood of newborns and 14-year-old children. Epigenetics 13, 655–664 (2018). sourceExperimental — accuracy not yet measured

In the paper

“Many of these studies were conducted by researchers who developed the biomarker and validated that specific biomarker across multiple cohorts, or researchers who used biomarkers developed by others and compared multiple biomarkers within one cohort, with differences between the two approaches illustrated in Figure 2.”

In the paper

“For example, different validation studies often account for distinct potential effect modifiers by controlling, adjusting, or stratifying for them.”

In the paper

“The criteria we defined for our overview resulted in exclusion of several studies.”

• Also stated in: § Extended Data · plain

  “As these strict inclusion criteria resulted in omission of numerous studies and biomarkers (see Extended Data Table 3 below for examples).”

In the paper

“Exclusion was primarily due to follow-up times below 10 years (Mahalanobis distance, 1; Cross-cohort evaluation of DNAm biomarkers,2), a lack of cross-validation in independent cohorts (miRNA biomarkers, 3; advanced biological aging predictor,4), a lack of evaluation of association with mortality (GlycanAge, 5; transcriptomic clock, 6), or a combination of the above (Plasma proteomic biomarker pan”

In the paper

“Many cohort studies provide alternative health outcome data (Table 3) which may support this goal, including data on specific chronic diseases, multi-morbidity, organ specific physiological integrity as well as physical and cognitive function.”

In the paper

“In studies validating blood-based biomarkers, most reported HRs for prediction of mortality risk are in the moderate range; however, a few studies have reported impressive metrics that render those biomarkers good potential candidates for use in preclinical and clinical studies.”

In the paper

“Nevertheless, several biomarkers of aging have been tested across multiple cohorts, with the most commonly-examined outcome being all-cause mortality.”

In the paper

“Additionally, studies can report performance metrics such as HRs in different ways (e.g., per standard deviation, compared to a reference group, or per unit increase), using different adjustment strategies for covariates (see Table 1).”

In the paper

“Beyond cohort-specific challenges, access to cohort data remains a general ongoing issue: applying for access to many government-funded datasets often requires lengthy paperwork and review processes, and can often span several months or years.”

In the paper

“As the costs of many omic assays are decreasing and cohorts/biobanks are increasingly incorporating multiple data modalities (Table 3), we expect multi-omic biomarkers to become common in the near future., emphasizing the need for accessible and standardized approaches.”

In the paper

“These factors are expected to affect the magnitude of the relationship between the omic biomarkers and aging-associated outcomes, representing another challenge for comparison of biomarkers in (and across) validation studies.”

In the paper

“Notably, aging biomarkers that are reproducible across population groups likely reflect fundamental mechanisms of aging biology.”

In the paper

“The translation of the science of aging to clinical applications holds substantial promises for the improvement of healthcare and the expansion of health expectancy, with the potential to both reduce health care expenditure and improve population health 75.”

What the sources say

• Not checked yet

  Checked before — its earlier verdict is held back until a new check runs.

  cites [75] Cohen, A. A. et al. Detection of a novel, integrative aging process suggests complex physiological integration. PLoS ONE 10, e0116489 (2015). source

In the paper

“An important prerequisite to accomplish this goal is the availability of solid and validated biomarkers of aging, which requires a process of validation to advance them into clinically valuable and actionable tools.”

In the paper

“However, cross-sectional age-associations can be biased by secular trends and selective attrition to study participation, which can preclude assessment of the predictive value of the marker in relation to future age-relevant outcomes.”

In the paper

“This issue is particularly important for proteomic assays that tend to generate many values ‘below the threshold of detection’, which may not be random but rather convey important information.”

In the paper

“Studies may also treat missing data or repeated measurements for biomarkers or outcomes differently, potentially influencing power or skewing performance estimates.”

In the paper

“For instance, in a population with high exposure to pollution or environmental contaminants, cancer biomarkers will appear to be highly predictive of all-cause mortality even if they are not in the general population.”

In the paper

“All included biomarkers/studies met the following inclusion criteria: 1) studies that propose new ways to quantify biological aging (predicting aging outcomes better than chronological age 27); 2) biomarkers measured entirely or primarily in blood; 3) biomarkers that are applicable to the general population and not specific to certain subpopulations with specific diseases; 4) biomarkers that”

What the sources say

• Not checked yetcites [27] Tian, Y. E. et al. Heterogeneous aging across multiple organ systems and prediction of chronic disease and mortality. Nat. Med. 29, 1221–1231 (2023). source