Aevum BioAI Cognitive Health and Longevity Scores: Scientific Rationale and Development of a Multidomain Biomarker Framework for Precision Healthy Ageing

 

Kris Ke Shyang See^1-6*, Bryan Yoke Jin Tay⁶, Scott Jun Kuang Low^1,5,6, Kyle Khang Lyn Tan^1,4,5,6, Wai Hong Cheang¹, Amarpreet Kaur Sarjit Singh¹, Kanimisha A/P Kumaragur^1,4, Nur Shamimah Binti Rahmadullah¹, Pavithra A/P Lakshmanan^1,4, Romel Mario Soyza¹, Nik Nassyiradina Putri Binti Nik Ahmat¹, Aisah binti Haji Mahit¹, Yin Ying Lim¹, Harshana A/P Ganesan¹, Miew Leng Khoo² and Ananyaa Sreekumar¹

Citation: See KKS, Tay BYJ, Low SJK, Tan KHL, Cheang WH, et al. Aevum BioAI Cognitive Health and Longevity Scores: Scientific Rationale and Development of a Multidomain Biomarker Framework for Precision Healthy Ageing. World AI J Med Healthc. 1(1):1-25.

Received: September 12, 2026 | Published: September 24 2026

Copyright^© 2026 Genesis Pub by See KKS, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0). This license permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are properly credited.

Abstract

Background: Population ageing is increasing the prevalence of cognitive impairment, cardiovascular disease, metabolic dysfunction, frailty and multimorbidity. Conventional healthcare is predominantly organised around the diagnosis and treatment of established disease, whereas contemporary geroscience increasingly emphasises earlier identification of biological vulnerability and preservation of healthspan. Ageing is multidimensional and involves interacting inflammatory, metabolic, vascular, nutritional, endocrine, neurological and lifestyle determinants. No single biomarker adequately captures these processes.

Objective: This review describes the scientific rationale, scoring architecture and proposed clinical applications of two complementary artificial-intelligence-assisted health-assessment systems developed by Aevum BioAI: the Aevum BioAI Cognitive Health Score (ABCHS) and the Aevum BioAI Longevity Score (ABLS).

Methods: A narrative review was conducted integrating evidence from geroscience, cognitive ageing, dementia prevention, cardiovascular medicine, metabolic health, biological-age research and lifestyle medicine. Particular attention was given to biomarkers incorporated into the current Aevum BioAI scoring algorithms. The scoring structure was derived from the current ABCHS and ABLS clinical reports.

Results: The ABCHS is a 40-point risk-stratification system, in which increasing scores represent increasing biological risk associated with future cognitive decline. It incorporates age, haematological variables, inflammation, renal and hepatic function, electrolytes, thyroid function, vitamin D, urinary beta-amyloid status and family history. The current risk bands are low (0–7), mild (8–15), moderate (16–23), high (24–31) and very high (32–40). The ABLS uses a complementary 100-point biological-resilience model, in which higher scores indicate more favourable biological and lifestyle profiles. Six domains contribute to the overall score: inflammation (20 points), metabolic health (20), organ function (20), nutrition and endocrine health (15), brain health (15), and lifestyle and clinical factors (10).

Conclusions: The Aevum BioAI framework represents a clinically accessible approach to integrating multidisystem biological and behavioral information into interpretable health scores. Its conceptual basis is consistent with geroscience, cardiovascular-health frameworks, multidomain dementia prevention and biological-age research. However, ABCHS and ABLS should currently be considered investigational risk-stratification and biological-health tools rather than diagnostic or prognostic instruments. Prospective longitudinal validation, external validation, calibration and comparison against established clinical models are required before individual disease or lifespan prediction can be claimed.

Keywords

Aevum BioAI; Artificial intelligence; Longevity; Cognitive health; Biomarkers; Biological ageing; Healthspan; Dementia prevention; Precision medicine; Risk stratification.

Introduction

Life expectancy has increased markedly over the past century, but improvements in lifespan have not necessarily produced proportional increases in healthy lifespan. Longer survival has therefore been accompanied by increasing prevalence of cardiovascular disease, diabetes, chronic kidney disease, neurodegeneration, frailty and multimorbidity.

The distinction between lifespan and healthspan is increasingly important. Lifespan represents the duration of life, whereas healthspan reflects the period during which an individual retains functional, cognitive and physiological capacity.

The World Health Organization conceptualizes healthy ageing around the maintenance of functional ability and intrinsic capacity rather than simply the absence of diagnosed disease [1[. Geroscience similarly proposes that apparently distinct chronic diseases frequently arise from shared biological mechanisms of ageing.

The expanded hallmarks of ageing describe 12 interacting processes: genomic instability, telomere attrition, epigenetic alteration, loss of proteostasis, disabled macro autophagy, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem-cell exhaustion, altered intercellular communication, chronic inflammation and dysbiosis [2].

These mechanisms operate across multiple organ systems. Consequently, ageing cannot be adequately represented by chronological age or a single laboratory measurement.

Cognitive decline illustrates this multidimensional biology particularly well. Dementia risk reflects interactions among age, genetic susceptibility, vascular disease, metabolic health, inflammation, sensory function and lifestyle. The 2024 Lancet Commission emphasized the potential importance of modifiable risk factors across the life course and reinforced the concept that dementia prevention begins considerably before overt cognitive impairment develops [3].

This creates a rationale for systems capable of combining multiple clinical signals into interpretable measures of health risk. The Aevum BioAI Cognitive Health Score (ABCHS) and Aevum BioAI Longevity Score (ABLS) were developed within this paradigm.

The two systems assess related but different dimensions:

• ABCHS: biological vulnerability associated with cognitive decline.
• ABLS: systemic biological resilience and modifiable determinants of healthy longevity.

 

Rather than attempting to replace clinical assessment, their intended function is to identify abnormalities, stratify biological risk, facilitate preventive intervention and monitor change longitudinally.

Conceptual Architecture of Aevum BioAI

The underlying premise of Aevum BioAI is that an individual's health trajectory is more informative than chronological age alone. Two individuals aged 60 years may have profoundly different inflammatory profiles, metabolic health, renal reserve, cardiovascular risk, physical activity, sleep quality and neurological biomarkers.

A multidimensional scoring model attempts to capture part of this heterogeneity. Importantly, the scoring directions of ABCHS and ABLS are opposite.

ABCHS

Higher score = greater biological cognitive-health risk.

Range = 0–40 points.

ABLS

Higher score = more favourable biological resilience.

Range = 0–100 points.

The ABCHS report describes the score as integrating routine laboratory biomarkers with inflammatory, metabolic and neurodegenerative markers while explicitly stating that it does not diagnose Alzheimer's disease or dementia.

Likewise, the ABLS report describes the longevity score as an integration of inflammatory, metabolic, organ-function, nutritional, brain-health and lifestyle biomarkers. Importantly, the report states that it estimates healthy ageing and biological resilience rather than predicting actual lifespan. This distinction is critical scientifically.

Aevum BioAI Cognitive Health Score

Scoring architecture

The current ABCHS contains ten biological domains:

Domain	Maximum risk score
Age	3
Haematology	8
Inflammation	3
Kidney function	3
Liver function	3
Electrolytes	2
Thyroid function	3
Vitamin D	3
Neurodegeneration	8
Family history	4
Maximum ABCHS	40

This is a particularly important aspect of the system because it connects measurement directly with intervention.

• Smoking: Smoking increases cardiovascular, oncological, respiratory and dementia risk [3,27].
• Blood pressure: Hypertension contributes substantially to stroke, cardiovascular disease, kidney disease and cognitive decline. Midlife hypertension is also recognised as a modifiable dementia-risk factor [3,28].
• Physical activity: Regular physical activity improves cardiorespiratory fitness, glucose regulation, cardiovascular health, muscle function and psychological wellbeing and is associated with reduced premature mortality [29].
• Sleep: Sleep duration and quality influence cardiometabolic function, immune regulation and cognitive health. Both inadequate and excessive sleep duration have been associated with adverse health outcomes [30,31].
• Alcohol: Alcohol-related risk is dose-dependent and varies by outcome. High or hazardous alcohol consumption increases hepatic, cardiovascular, cancer and neurological disease burden [32].
• Stress: Chronic psychological stress may influence autonomic regulation, inflammation, sleep and cardiovascular health [33].

 

ABLS Categories

The current scoring system divides ABLS into six categories:

ABLS	Classification
90–100	Exceptional Longevity Profile
80–89	Healthy Longevity
70–79	Good – Optimisation Recommended
60–69	Moderate Risk
50–59	High Risk
0–49	Very High Risk

 

Relationship With Life's Essential 8

The American Heart Association's Life's Essential 8 provides an important precedent for composite positive-health scoring [25].

It combines:

• diet;
• physical activity;
• nicotine exposure;
• sleep;
• body mass index;
• blood lipids;
• glucose; and
• blood pressure.

 

The framework uses a 0–100 cardiovascular-health scale. ABLS extends a similar philosophy beyond cardiovascular health by incorporating inflammation, renal and hepatic function, nutritional/endocrine variables and brain-health indicators.

Relationship With Biological-Age Algorithms

Biological-age models attempt to quantify heterogeneity in ageing beyond chronological age. Horvath demonstrated that DNA methylation signatures can estimate age across tissues [35].

Levine and colleagues subsequently developed Phenotypic Age, linking clinical biomarkers with morbidity, healthspan and mortality before translating this phenotype into an epigenetic clock [36].

GrimAge improved mortality and healthspan prediction using methylation surrogates associated with smoking and plasma proteins [37].

DunedinPACE was developed to estimate the pace of biological ageing rather than simply biological age [38]. Aevum BioAI differs conceptually from these approaches.

ABLS is not simply intended to answer:

"How biologically old is this person?"

Instead, it attempts to answer:

"Which biological systems appear resilient, which are deteriorating and what potentially modifiable factors are driving the difference?"

This distinction may make ABLS particularly appropriate for preventive clinical practice.

Longitudinal Scoring May Be More Important Than a Single Score

A single biomarker result represents only one point in time. The more clinically important variable may be change.

For example: ABLS 58 → 66 → 76 → 84

following intervention could indicate improvement in several underlying systems.

Similarly: ABCHS 5 → 8 → 13 → 17

Could indicate progressive accumulation of risk even before subjective cognitive symptoms appear.

The current Aevum reports already support this longitudinal philosophy.

ABCHS recommends repeat assessment approximately every 12 months.

ABLS recommends repeat assessment approximately every 6–12 months.

Proposed Aevum BioAI Longitudinal Model

Future risk could conceptually be represented as:

Hi,t=f(Bi,t,Mi,t,Oi,t,Ni,t,Ci,t,Li,t,Fi)

where:

• H = overall health trajectory,
• B = inflammatory/haematological biology,
• M = metabolic health,
• = organ function,
• N = nutritional/endocrine state,
• C = cognitive/neurological markers,
• L = lifestyle,
• F = familial/genetic susceptibility,
• i = individual,
• and t = time.

The clinically interesting quantity may ultimately be:

ΔH=Ht2−Ht1

rather than H at a single assessment.

Role of Artificial Intelligence

The current score architecture can operate through explicit predefined rules.

However, accumulation of sufficiently large prospective datasets would allow Aevum BioAI to progress toward genuine data-driven prediction.

Potential machine-learning techniques could identify:

• nonlinear relationships;
• biomarker interactions;
• age-specific effects;
• sex-specific associations;
• temporal trajectories;
• clusters of risk;
• differential responses to intervention.

Importantly, AI should not automatically replace the transparent score.

A clinically preferable structure may be:

• Layer 1: Transparent score. Physicians can see exactly which variables contribute points.
• Layer 2: AI risk model. Machine learning analyses interactions among variables.
• Layer 3: Longitudinal model. The platform evaluates direction and rate of change.
• Layer 4: Clinical decision support Interventions are suggested based upon modifiable drivers.

Such an architecture preserves clinical interpretability while taking advantage of AI.

Development of the Scoring Algorithm

For a formal methods publication, ABCHS and ABLS should be described as weighted additive composite scores.

For ABCHS: ABCHS=∑j=110Rj

Where Rj represents the risk contribution from each domain and: 0≤ABCHS≤40

For ABLS: ABLS=∑k=16Pk

Where Pk represents points retained within each biological-resilience domain and: 0≤ABLS≤100

The full scoring matrix should eventually be published as supplementary material.

Example of the required ABCHS supplement.

Marker	Result range	Risk points
hs-CRP	Optimal range	0
	threshold 2	x
	threshold 3	x
	highest-risk range	3
Vitamin D	Optimal range	0
	intermediate	1
	intermediate	2
	severe abnormality	3

 

This produces a clinically intuitive sequence:

Measure → Stratify → Identify → Intervene → Reassess

The system therefore shifts emphasis from static laboratory reporting toward longitudinal health optimisation.

Proposed Validation Strategy

The most important next step is prospective validation. A multicenter longitudinal cohort would be preferable.

Study population

• Ideally: n = 1,000–5,000 participants
• Age: 40–80 years
• Follow-up: minimum 3 years; preferably 5–10 years.

 

Recruitment should include participants from different ethnic and socioeconomic backgrounds. A Malaysian cohort would be particularly valuable because most biological-age and dementia-risk algorithms have historically been developed in Western populations.

 

ABCHS Validation Endpoints

Baseline evaluation should include:

• ABCHS;
• MoCA;
• MMSE where appropriate;
• subjective cognitive complaints;
• medical history;
• medications;
• depression assessment;
• sleep assessment;
• vascular-risk factors.

A nested subgroup could undergo:

• plasma p-tau217;
• p-tau181;
• Aβ42/40;
• NfL;
• GFAP;
• MRI;
• amyloid PET where clinically justified.

Primary endpoints could include:

• change in MoCA
• incident mild cognitive impairment
• incident dementia
• neurological biomarker progression
• structural brain changes.

 

ABLS Validation Endpoints

Potential outcomes include:

• cardiovascular events;
• diabetes;
• chronic kidney disease;
• cancer;
• frailty;
• multimorbidity;
• hospitalisation;
• functional impairment;
• major adverse health events;
• all-cause mortality.

Healthspan outcomes may ultimately be more scientifically relevant than crude lifespan.

Statistical Validation

A rigorous validation programme should evaluate four dimensions.

• Discrimination: Receiver-operating-characteristic curves and area under the curve.
• Calibration: Agreement between predicted and observed outcomes.
• Reclassification: Whether ABCHS or ABLS adds meaningful information beyond conventional models.
• Clinical utility: Decision-curve analysis to determine whether using the score improves clinical decision-making.

Other analyses should include:

• sensitivity;
• specificity;
• PPV;
• NPV;
• calibration slope;
• Brier score;
• net reclassification improvement;
• test-retest reliability;
• inter-laboratory reproducibility.

Comparison Against Existing Models

ABCHS should eventually be compared with established dementia-risk frameworks and cognitive assessment.

ABLS should be compared with:

• chronological age;
• Life's Essential 8;
• frailty index;
• Phenotypic Age;
• conventional cardiovascular-risk scores;
• metabolic-health indices.

Demonstrating correlation is insufficient. The key question is whether Aevum provides incremental predictive value.

Avoiding Algorithmic Bias

AI-based health systems require careful evaluation for bias. Age, sex, ethnicity, socioeconomic status, laboratory methodology and underlying disease prevalence can affect biomarker distributions. Models trained predominantly on one population may perform poorly in another.

Development should therefore include:

• sex-stratified analysis;
• age-stratified analysis;
• ethnicity-stratified validation;
• missing-data analysis;
• calibration across populations;
• external replication.

Explainability should remain central. Clinicians should be able to identify which variables caused an individual's score.

Future Biomarker Expansion

Future ABCHS and ABLS versions could evaluate additional biomarkers, but expansion should be evidence-driven.

Cardiometabolic

• ApoB
• lipoprotein(a)
• fasting insulin
• HOMA-IR
• uric acid

 

Inflammation

• IL-6
• TNF-α
• GlycA

 

Neurodegeneration

• p-tau217
• p-tau181
• NfL
• GFAP
• plasma Aβ42/40

 

Cardiovascular function

• pulse-wave velocity
• coronary calcium
• cardiorespiratory fitness

 

Biological ageing

• epigenetic clocks
• proteomic age
• metabolomic age
• glycomic age

 

Wearables

• heart-rate variability
• sleep efficiency
• sleep regularity
• resting heart rate
• physical activity
• gait
• continuous glucose profiles

 

The criterion for inclusion should be:

Does the new biomarker materially improve prediction, intervention selection or monitoring?

Not simply whether it is technologically novel.

Toward the Aevum Digital Health Twin

A longer-term objective could be transformation of ABCHS and ABLS into a dynamic individual health model.

A person's record could integrate:

ABCHS+ABLS+genomics+imaging+wearables+longitudinal biomarkers

A continuously updated model could then estimate:

• current biological resilience
• trajectory of ageing
• major risk drivers
• response to previous interventions
• potentially modifiable targets.

This can be conceptualised as an Aevum Digital Health Twin.

Rather than asking:

"What is my biological age?"

precision longevity medicine could ask: "Which parts of my biology are ageing unfavorably, how rapidly are they changing, and which interventions are most likely to alter that trajectory?"

That is arguably the more clinically meaningful application of BioAI.

Limitations

Several limitations of the current framework must be clearly acknowledged.

• First, the score weights and categories presently represent a proposed clinical algorithm, not coefficients derived from a published perspective cohort.
• Second, several included biomarkers are nonspecific.
• Third, individual biomarker results can be affected by transient illness, medication, hydration, laboratory methodology and environmental exposures.
• Fourth, the lifestyle domain partly depends upon self-reported information.
• Fifth, the urinary beta-amyloid component requires substantially greater validation.
• Sixth, the biological-age and lifespan projections within ABLS should not be interpreted as individual actuarial predictions.

 

The current report appropriately acknowledges that the biological-age calculation has not yet been validated against longitudinal disease, frailty, hospitalisation or mortality outcomes.  These limitations should be described transparently because doing so strengthens rather than weakens the scientific credibility of the programme.

Conclusion

The Aevum BioAI Cognitive Health Score and Aevum BioAI Longevity Score represent complementary attempts to translate complex biological information into practical measures of health trajectory.

ABCHS is a 40-point cognitive-health risk score incorporating age, haematology, inflammation, renal and hepatic function, electrolytes, thyroid status, vitamin D, neurodegenerative markers and family history.

ABLS is a 100-point biological-resilience score incorporating inflammation, metabolic health, organ function, nutrition and endocrine health, brain health, and lifestyle-clinical factors.

Their scientific premise is supported by an increasingly coherent body of evidence showing that ageing and cognitive decline result from interconnected biological and behavioral processes rather than isolated abnormalities.

The principal innovation of Aevum BioAI is therefore not an individual laboratory test.

It is the attempt to integrate:  biomarkers + organ function + brain health + lifestyle + longitudinal change

At present, both systems should remain investigational and should complement rather than replace clinical judgement. Prospective validation will be necessary to determine their calibration, discrimination, reproducibility and incremental clinical value.

If those studies demonstrate robust predictive performance, Aevum BioAI could evolve from a biomarker scoring platform into a longitudinal precision-health system capable of detecting biological deterioration before the development of established disease. The goal of longevity medicine need not be to predict the exact age at which an individual will die.

A more clinically useful objective is to determine: where biological vulnerability is developing, whether the trajectory is worsening, and whether intervention can alter it before irreversible disease occurs.

That represents the central proposition underlying Aevum BioAI.

References

1. World Health Organization. (2025) World Report on Ageing and Health. Geneva: WHO.
2. López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. (2023) Hallmarks of aging: an expanding universe. Cell. 186(2):243-78.
3. Livingston G, Huntley J, Liu KY, et al. (2024) Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. Lancet. 404(10452):572-628.
4. Kennedy BK, Berger SL, Brunet A, et al. (2014) Geroscience: linking aging to chronic disease. Cell. 159(4):709-13.
5. Wolters FJ, Zonneveld HI, Licher S, et al. (2019) Hemoglobin and anemia in relation to dementia risk and accompanying changes on brain MRI. Neurology. 93:e917–e926.
6. Andro M, Le Squere P, Estivin S, Gentric A. (2013) Anaemia and cognitive performances in the elderly: a systematic review. Eur J Neurol. 20:1234-40.
7. Patel KV, Ferrucci L, Ershler WB, Longo DL, Guralnik JM. (2009) Red blood cell distribution width and the risk of death in middle-aged and older adults. Arch Intern Med. 169:515-23.
8. Mohammadi A, Mohammadi M, Almasi-Dooghaee M, Mirmosayyeb O. (2024) Neutrophil to lymphocyte ratio in Alzheimer's disease: a systematic review and meta-analysis. PLoS One. 19:e0305322.
9. Darweesh SKL, Wolters FJ, Ikram MA, de Wolf F, Bos D, Hofman A. (2018) Inflammatory markers and the risk of dementia and Alzheimer's disease: a meta-analysis. Alzheimer's Dement. 14:1450–1459.
10. Kuring JK, Mathias JL, Ward L, Tachas G. (2026) Inflammatory markers associated with dementia: a systematic review and meta-analysis. J Psychiatr Res. 197:26–40.
11. Pei X, Bakerally NB, Wang Z, et al. (2025) Kidney function and cognitive impairment: a systematic review and meta-analysis. Ren Fail. 47:2463565.
12. Maung ST, Hemrungrojn S, Chaiteerakij R. (2026) Cognitive impairment in metabolic dysfunction-associated steatotic liver disease: a systematic review and meta-analysis. Diabetes Metab Syndr. 20:103467.
13. Liamis G, Milionis HJ, Elisaf M. Hyponatremia in patients with infectious diseases. J Infect. 2011;63:327–335.
14. Salehipour A, Dolatshahi M, Haghshomar M, Amin J. The role of thyroid dysfunction in Alzheimer's disease: a systematic review and meta-analysis. J Prev Alzheimers Dis. 2023;10:276–286.
15. Zhang XX, Wang HR, Wei M, et al. Association of vitamin D levels with risk of cognitive impairment and dementia: a systematic review and meta-analysis of prospective studies. J Alzheimers Dis. 2024;98:373–385.
16. Chen WY, Cheng YC, Chiu CC, et al. Effects of vitamin D supplementation on cognitive outcomes: a systematic review and meta-analysis. Neuropsychol Rev. 2024;34:568–580.
17. Jack CR Jr, Bennett DA, Blennow K, et al. NIA-AA Research Framework: toward a biological definition of Alzheimer's disease. Alzheimer's Dement. 2018;14:535–562.
18. Hansson O. Biomarkers for neurodegenerative diseases. Nat Med. 2021;27:954–963.
19. Teunissen CE, Verberk IMW, Thijssen EH, et al. Blood-based biomarkers for Alzheimer's disease: towards clinical implementation. Lancet Neurol. 2022;21:66–77.
20. Ashton NJ, Brum WS, Di Molfetta G, et al. Diagnostic accuracy of a plasma phosphorylated tau 217 immunoassay for Alzheimer disease pathology. JAMA Neurol. 2024;81:255–263.
21. Belloy ME, Napolioni V, Greicius MD. A quarter century of APOE and Alzheimer's disease: progress to date and the path forward. Neuron. 2019;101:820–838.
22. Franceschi C, Garagnani P, Parini P, Giuliani C, Santoro A. Inflammaging: a new immune-metabolic viewpoint for age-related diseases. Nat Rev Endocrinol. 2018;14:576–590.
23. Emerging Risk Factors Collaboration. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease. Lancet. 2010;375:2215–2222.
24. Sarwar N, Gao P, Seshasai SRK, et al. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease. Lancet. 2010;375:2215–2222.
25. Lloyd-Jones DM, Allen NB, Anderson CAM, et al. Life's Essential 8: updating and enhancing the American Heart Association's construct of cardiovascular health. Circulation. 2022;146:e18–e43.
26. Stabler SP. Vitamin B12 deficiency. N Engl J Med. 2013;368:149–160.
27. Zhong G, Wang Y, Zhang Y, Guo JJ, Zhao Y. Smoking is associated with an increased risk of dementia: a meta-analysis of prospective cohort studies. PLoS One. 2015;10:e0118333.
28. Williamson JD, Pajewski NM, Auchus AP, et al. Effect of intensive vs standard blood pressure control on probable dementia. JAMA. 2019;321:553–561.
29. Ekelund U, Tarp J, Steene-Johannessen J, et al. Dose-response associations between accelerometry measured physical activity and sedentary time and all-cause mortality. BMJ. 2019;366:l4570.
30. Itani O, Jike M, Watanabe N, Kaneita Y. Short sleep duration and health outcomes: a systematic review, meta-analysis and meta-regression. Sleep Med. 2017;32:246–256.
31. Sabia S, Fayosse A, Dumurgier J, et al. Association of sleep duration in middle and old age with incidence of dementia. Nat Commun. 2021;12:2289.
32. GBD 2016 Alcohol Collaborators. Alcohol use and burden for 195 countries and territories, 1990–2016. Lancet. 2018;392:1015–1035.
33. Cohen S, Janicki-Deverts D, Miller GE. Psychological stress and disease. JAMA. 2007;298:1685–1687.
34. Ngandu T, Lehtisalo J, Solomon A, et al. A 2-year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER). Lancet. 2015;385:2255–2263.
35. Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14:R115.
36. Levine ME, Lu AT, Quach A, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging. 2018;10:573–591. doi:10.18632/aging.101414.
37. Lu AT, Quach A, Wilson JG, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging. 2019;11:303–327.
38. Belsky DW, Caspi A, Corcoran DL, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022;11:e73420.
39. Kivipelto M, Mangialasche F, Ngandu T. Lifestyle interventions to prevent cognitive impairment, dementia and Alzheimer disease. Nat Rev Neurol. 2018;14:653–666.
40. Peters R, Booth A, Rockwood K, Peters J, D'Este C, Anstey KJ. Combining modifiable risk factors and risk of dementia: a systematic review and meta-analysis. BMJ Open. 2019;9:e022846.
41. Sofi F, Abbate R, Gensini GF, Casini A. Accruing evidence on benefits of adherence to the Mediterranean diet on health. BMJ. 2010;341:c5684.
42. Lourida I, Soni M, Thompson-Coon J, et al. Mediterranean diet, cognitive function, and dementia: a systematic review. Epidemiology. 2013;24:479–489.
43. Bautmans I, Knoop V, Amuthavalli Thiyagarajan J, et al. WHO working definition of vitality capacity for healthy longevity monitoring. Lancet Healthy Longev. 2022;3:e789–e796.
44. Beard JR, Officer A, de Carvalho IA, et al. The World report on ageing and health: a policy framework for healthy ageing. Lancet. 2016;387:2145–2154.
45. Ferrucci L, Gonzalez-Freire M, Fabbri E, et al. Measuring biological aging in humans: a quest. Aging Cell. 2020;19:e13080.
46. Justice JN, Ferrucci L, Newman AB, et al. A framework for selection of blood-based biomarkers for geroscience-guided clinical trials. Geroscience. 2018;40:419–436.
47. Sebastiani P, Thyagarajan B, Sun F, et al. Biomarker signatures of aging. Aging Cell. 2017;16:329–338.
48. Cohen AA, Milot E, Yong J, et al. A novel statistical approach shows evidence for multi-system physiological dysregulation during aging. Mech Ageing Dev. 2013;134:110–117.