The Consensus 14: A New Standard for Biological Age

The End of the “Guesswork Era”

For decades, longevity optimisation was a decentralised exercise in inference — measuring surrogate endpoints, extrapolating from disease populations, and hoping that interventions validated in sick patients translated to healthy ones.

In late 2024, an international panel of 460 aging researchers conducted a three-round structured Delphi consensus study — the most rigorous methodology available for achieving expert consensus when empirical data is incomplete — and published the results in The Journals of Gerontology: Series A (Perri et al., December 2024, in print May 2025). The outcome: formal consensus on 14 biomarkers of aging suitable for use as outcome measures in human intervention studies. Consensus was defined as ≥70% agreement among panel members. All 14 markers exceeded this threshold, with agreement ranging from 70% to 98%.

At the Forge, the Consensus 14 are the only markers that determine whether a protocol is working. Every individual briefing in this research library maps to one or more of these 14 domains. Every intervention is evaluated against its capacity to shift these specific measurable outputs.

Forge Editorial Note — Accurate Framing: The Perri et al. consensus paper explicitly states that these are “potential biomarkers of aging which may be used as outcome measures in intervention studies” and that “future aging research should identify which combination has the greatest utility.” The Forge adopts this list as its primary tracking framework — and has developed Forge-specific optimal targets for each marker based on the broader longevity literature. Those targets represent the Forge’s best-evidence position; they are not directly stated in the Perri et al. paper, which establishes the framework rather than individual optimal values.

I. The Physiological & Inflammatory Panel — The Internal Climate

These four blood-derived markers monitor the somatotropic axis and the systemic inflammatory environment. They represent the metabolic and inflammatory friction operating within the system — the upstream drivers that most of the Forge Pillar 01 and Pillar 04 protocols are designed to address.

• IGF-1: Measures the activity of the growth hormone / insulin-like growth factor axis and nutrient-sensing pathway tone. The U-shaped mortality relationship — both floor and ceiling carrying risk — is documented across n>30,000 in the landmark meta-analysis. See: IGF-1 Briefing.

• GDF-15 (Growth Differentiation Factor-15): A mitokine — a stress-induced cytokine secreted by cells undergoing mitochondrial dysfunction, DNA damage, or inflammatory activation. GDF-15 is one of the most sensitive available blood markers for systemic biological stress, rising earlier than most conventional markers in response to cellular damage. It is not yet as widely available as hs-CRP or IL-6 on standard clinical panels, but its inclusion in the Consensus 14 signals its growing clinical relevance.

• hs-CRP: The gold standard clinical marker for systemic low-grade inflammation — the downstream reporter of IL-6 and the primary Pillar 04 tracking metric. In the Emerging Risk Factors Collaboration (n=160,309), each SD increase predicted CV mortality with RR=1.55. See: hs-CRP Briefing.

• IL-6: The upstream cytokine orchestrating the chronic inflammatory cascade that generates CRP, drives sarcopenia, and accelerates white matter deterioration. The MESA cohort (n=6,622) confirmed IL-6 predicts all-cause mortality HR=1.98 for highest vs. lowest tercile, independently of hs-CRP. See: IL-6 Briefing.

II. The Functional Pillars — The Hardware Reserve

The consensus panel identified nine functional and physical biomarkers as outcome measures. This is the largest single domain in the Consensus 14 — reflecting the field’s recognition that longevity is measured not in bloodwork alone, but in the preserved capacity to live functionally. These markers map primarily to Pillar 02: Physical Architecture.

1. Muscle Mass (SMMI): Skeletal muscle as the primary metabolic glucose disposal sink and amino acid reserve. Wang et al. meta-analysis (n=81,358): low SMMI associated with RR=1.57 for all-cause mortality. See: Muscle Mass Index Briefing.

2. Muscle Strength: Maximal force output as an independent longevity signal, distinct from mass. García-Hermoso meta-analysis (~2 million participants): higher muscular strength independently associated with lower all-cause and cancer mortality.

3. Hand Grip Strength: The single most practical and accessible proxy for total musculoskeletal and neuromuscular integrity. PURE study (n=139,691, 17 countries): HR=1.16 per 5 kg grip reduction, stronger predictor than systolic blood pressure. Wu et al. meta-analysis (n=3,002,203): all-cause mortality HR=1.41 lowest vs. highest category. See: Grip Strength Briefing. In the BASE-II validation (Vetter et al., 2025), hand grip strength was among the five Consensus 14 markers independently associated with mortality after full adjustment.

4. Gait Speed: Walking speed under normal or dual-task conditions — now recognised as a vital sign reflecting neurological integrity, lower-body power, cardiovascular reserve, and CNS-musculoskeletal coordination. In multiple meta-analyses, gait speed below 0.8 m/s is associated with substantially elevated mortality risk and loss of functional independence.

5. Standing Balance: Quantifies the integration of visual, vestibular, and proprioceptive systems — increasingly used as a proxy for overall nervous system integrity in ageing populations. Single-leg balance time below 10 seconds at any age is a well-validated frailty predictor.

6. Timed-Up-and-Go (TUG): A compound assessment of functional mobility, requiring the coordinated sequence of rising from a chair, walking 3 metres, turning, and returning. TUG time > 12 seconds is a validated indicator of fall risk and functional decline. It integrates lower-body strength, balance, gait, and executive function simultaneously.

7. Frailty Index: A composite algorithmic score measuring cumulative biological reserve — typically computed as the count of health deficits present (symptoms, signs, diagnoses, functional limitations) divided by the total assessed. Higher frailty indices predict hospitalisation, disability onset, and mortality independently of any individual marker. The Forge uses frailty index as the contextual integrator for the functional panel — it captures what the individual markers collectively express.

8. Cognitive Health: High-resolution testing of processing speed, executive function, and working memory — monitoring Pillar 03 trajectory. As established in the Cognitive Processing Speed Briefing, reaction time (UK Biobank n=54,019) adds independent mortality information beyond fitness or intelligence. In the BASE-II validation, cognitive health was one of the five markers independently associated with mortality after full covariate adjustment.

9. Blood Pressure: A critical hemodynamic proxy for vascular health — systolic and diastolic BP remaining among the most practically accessible cardiovascular risk predictors. The Forge’s preferred direct measure is cfPWV (arterial stiffness), with BP as the accessible daily proxy. See: Vascular Age Briefing.

III. The Epigenetic “Speedometer”

The consensus panel identified DNA methylation/epigenetic clocks as the single epigenetic domain in the Consensus 14 — representing the molecular substrate underlying biological age accumulation and aging velocity. Within this category, the Forge employs DunedinPACE as the primary instrument, for reasons established in the DunedinPACE Briefing:

• Unlike odometer-style clocks (GrimAge, PhenoAge, Horvath), DunedinPACE measures aging velocity rather than accumulated mileage
• It is the only epigenetic clock with direct RCT evidence of intervention-driven slowing (CALERIE trial, n=220, 2–3% reduction from 25% caloric restriction over 2 years)
• In the BASE-II independent validation (Vetter et al., medRxiv, November 2025, n=1,083, mean 7.4-year follow-up), DunedinPACE emerged as the strongest single predictor of all-cause mortality among all 14 Consensus markers in adjusted models — confirming its position as the Forge’s primary KPI

Forge Verdict: The objective is to maintain DunedinPACE consistently below 1.0 on annual testing. A score below 0.95 places you in the top ~25% of the population for aging deceleration. A sustained trajectory of declining DunedinPACE across three or more annual tests — regardless of the absolute starting value — is the primary evidence that the Forge protocol is working. For contextual framing of what lifestyle intervention can realistically achieve, the CALERIE data (2–3% slowing from intensive caloric restriction) is the best available human benchmark.

What the Consensus 14 Is — and Is Not

Understanding what the Perri et al. paper actually established prevents both overclaiming and underclaiming.

What it is: A rigorous, transparent, internationally validated framework identifying the 14 biomarkers most suitable for measuring whether aging interventions are working in human studies. The Delphi methodology — 460 invited experts, three rounds of structured voting, ≥70% consensus threshold — is among the most robust available for generating scientific consensus in fields where RCT data on outcomes is decades away.

What it is not: A complete clinical protocol with validated individual optimal targets, a prediction of which combination of markers provides maximum prognostic utility, or a claim that tracking these 14 markers is sufficient to fully characterise biological age. The paper explicitly calls for “future research to identify which combination has the greatest utility” — the BASE-II validation paper (Vetter et al.) is the first step in answering that question.

The Forge application: Every article in this research library addresses one or more Consensus 14 markers. Every Forge Optimal target cited in those briefings is drawn from the deepest available individual-marker literature — the specific papers and cohort data most relevant to each marker’s optimal range. The Consensus 14 framework provides the architecture; the individual briefings provide the evidence-grounded targets within each domain.

How to Use the Roadmap

If you are not tracking at least 8 of the 14 markers with serial data — not just single-point readings — you are operating without a trajectory. The Forge Optimal Ranges for each marker are detailed in the individual Pillar briefings. Use them as your targets. Use DunedinPACE as the integrator that tells you whether the individual marker improvements are collectively slowing your biological clock.

The combination of markers that most consistently predicts mortality in the BASE-II validation data: DunedinPACE, IL-6, Cognitive Health, Hand Grip Strength, and Standing Balance. If you can only track five of the 14 markers, these five — spanning epigenetic, inflammatory, neural, and functional domains — provide the highest-information-density cross-system biological age readout currently available.

References

• Perri G. et al. (60 expert authors), The Journals of Gerontology: Series A (2025): “An Expert Consensus Statement on Biomarkers of Aging for Use in Intervention Studies.” 3-round Delphi study, 460 invited panelists, 60 completers. ≥70% consensus on 14 biomarkers spanning physiological (IGF-1, GDF-15), inflammatory (hs-CRP, IL-6), functional (muscle mass, muscle strength, HGS, TUG, gait speed, standing balance, frailty index, cognitive health, blood pressure), and epigenetic (DNA methylation/epigenetic clocks) domains. Published December 2024 (online first); Vol. 80, No. 5, May 2025. DOI: 10.1093/gerona/glae297
• Vetter V.M. et al., medRxiv (November 2025): “Comparing Fourteen Consensus Biomarkers of Aging: Epigenetic Pace of Aging as the Strongest Predictor of Mortality in BASE-II.” Berlin Aging Study II, n=1,083, mean follow-up 7.4 years (range 3.9–10.4). DunedinPACE: strongest adjusted predictor of all-cause mortality among all 14 markers. IL-6, cognitive health, HGS, muscle strength, and standing balance also independently significant after full adjustment (age, sex, lifestyle, genetic ancestry). DOI: 10.1101/2025.11.20.25340648
• Belsky D.W. et al., eLife (2022): DunedinPACE development and validation. n=1,037, Dunedin Study ages 26–45. ICC=0.96 test-retest reliability; validated in 5 external datasets. DOI: 10.7554/eLife.73420
• Waziry R. et al., Nature Aging (2023): CALERIE RCT, n=220. 25% CR: 2–3% DunedinPACE slowing; PhenoAge and GrimAge non-significant. Upper-bound benchmark for lifestyle-only DunedinPACE intervention. DOI: 10.1038/s43587-022-00357-y
• Individual Forge Pillar Briefings: IGF-1, hs-CRP, IL-6, Fasting Insulin, HbA1c, ApoB, Cystatin C, HRV Trends, Deep Sleep %, Cognitive Processing Speed, Grip Strength, Muscle Mass Index, Vascular Age (cfPWV), Microbiome Diversity, DNA Methylation, DunedinPACE — each providing evidence-grounded Forge Optimal targets within their respective Consensus 14 domain.