A proposed panel of biomarkers of healthy ageing

This was not an intervention study but a systematic review and expert consensus process to answer: *Which biomarkers best capture the multi-system decline of healthy ageing, and which specific tests should researchers use to measure them?*

The authors searched for biomarkers across five ageing-related domains:

**Physical capability:** Strength, locomotion, balance, dexterity

**Cognitive function:** Memory, processing speed, executive function

**Physiological function:** Cardiovascular function, lung function, glucose metabolism, musculoskeletal function

**Endocrine function:** Hypothalamic-pituitary-adrenal (HPA) axis, sex hormones, growth hormones

**Immune function:** Inflammatory markers

They compared candidate biomarkers against criteria: the biomarker must change with age in a predictable direction, be measurable in large populations, have established reference ranges, and show associations with health outcomes like mortality or disability.

The outcome was a recommended panel of specific tests and biomarkers, not a single number. No comparator group was used because this was a measurement-development study, not a trial.

This was a meta-analysis and expert consensus, so there was no single study population. The authors drew on:

Existing systematic reviews and meta-analyses of ageing biomarkers (number of studies reviewed not explicitly stated, but the search covered multiple databases with no date restrictions up to 2012)

The NIH Toolbox for assessment of neurological and behavioural function (normed on ~4,859 adults aged 3–85 in the US)

Expert input from an international workshop held in Newcastle, UK, on 22–23 October 2012, with "international experts" (number not specified) from ageing research, epidemiology, and clinical medicine

The target population for the biomarker panel is community-dwelling older adults (typically aged 60+), but the biomarkers are applicable across adulthood from age ~40 onward, when age-related decline becomes measurable.

The authors did not collect new data. Instead, they synthesised existing measurement instruments and recommended specific tests for each biomarker domain:

**Physical capability:**

Grip strength: measured with a hand dynamometer (Jamar or equivalent), in kilograms, best of two trials per hand

Gait speed: usual pace over 4–6 metres, in metres per second

Chair rise time: time to stand from a chair five times without using arms, in seconds

Balance: ability to stand on one leg for 30 seconds (eyes open), timed in seconds

Dexterity: Purdue Pegboard test (number of pegs placed in 30 seconds)

**Cognitive function:**

Memory: Rey Auditory Verbal Learning Test (RAVLT) or NIH Toolbox Picture Sequence Memory Test (score range depends on version)

Processing speed: Digit Symbol Substitution Test (DSST) from the Wechsler Adult Intelligence Scale (number of correct symbols in 90–120 seconds)

Executive function: Trail Making Test Part B (time to complete, in seconds; lower = better) or NIH Toolbox Dimensional Change Card Sort Test

**Physiological function:**

Cardiovascular: systolic and diastolic blood pressure (mmHg), resting heart rate (bpm)

Lung function: forced expiratory volume in 1 second (FEV1, in litres) and forced vital capacity (FVC, in litres), measured by spirometry

Glucose metabolism: fasting glucose (mmol/L or mg/dL) and HbA1c (%)

Musculoskeletal: bone mineral density (BMD) by DXA scan (T-score), body mass index (BMI, kg/m²)

**Endocrine function:**

HPA axis: morning cortisol (nmol/L or μg/dL), typically from blood or saliva

Sex hormones: total testosterone (nmol/L) in men; oestradiol (pmol/L) in women

Growth hormones: insulin-like growth factor 1 (IGF-1, ng/mL)

**Immune function:**

Inflammatory markers: C-reactive protein (CRP, mg/L), interleukin-6 (IL-6, pg/mL), tumour necrosis factor-alpha (TNF-α, pg/mL)

**Study design:** This was a systematic review with expert consensus, not a meta-analysis in the traditional sense (no pooled effect sizes). The authors conducted comprehensive literature searches across multiple databases (PubMed, Web of Science, PsycINFO) for reviews and primary studies on biomarkers of ageing, then convened an expert workshop to refine recommendations.

**Selection process:** The authors identified candidate biomarkers from existing systematic reviews and meta-analyses where available. For domains lacking reviews (e.g., dexterity), they consulted the NIH Toolbox and expert opinion. They applied explicit criteria: the biomarker must (1) show consistent age-related change, (2) be measurable in field settings, (3) have established normative data, and (4) predict health outcomes like mortality, disability, or hospitalisation.

**Expert consensus:** A two-day workshop in Newcastle (October 2012) brought together international experts who critiqued draft recommendations. The final panel reflects majority agreement, but the paper does not report formal consensus methods (e.g., Delphi process) or quantify agreement levels.

**What this design can prove:** This process can identify which biomarkers have the strongest evidence base for tracking ageing. It provides a standardised measurement framework that allows comparison across studies. The panel is evidence-based and expert-vetted.

**What this design cannot prove:** This is not a prospective study. It cannot tell you which biomarkers are *causal* drivers of ageing versus mere correlates. It cannot tell you how fast these biomarkers change per year in a given individual (though other studies provide those rates). It cannot tell you whether intervening on one biomarker (e.g., lowering CRP) will slow ageing in other domains. The panel is a measurement toolkit, not a validated composite score.

**Major methodological weaknesses:**

No formal systematic review protocol was registered (e.g., PROSPERO)

The search strategy and inclusion/exclusion criteria are not fully detailed

Expert selection and consensus methods are opaque — we don't know who the experts were, how many participated, or whether there was disagreement

The panel was not validated against a gold standard (because none exists) or tested prospectively in a new cohort

Some biomarkers (e.g., dexterity, balance) have weaker evidence than others (e.g., grip strength, gait speed)

The panel is biased toward biomarkers measurable in field settings, potentially missing more invasive but informative markers (e.g., telomere length, epigenetic clocks, which were not included)

The authors did not report quantitative effect sizes (no meta-analytic pooling). Instead, they present a narrative synthesis of which biomarkers show consistent age-related decline and which specific tests are recommended. Key findings by domain:

**Physical capability:**

Grip strength declines ~0.5–1.0 kg per year after age 50 in men, ~0.3–0.6 kg per year in women (based on multiple cohort studies)

Gait speed declines ~0.01–0.02 m/s per year after age 60

Chair rise time increases ~0.5–1.0 seconds per decade

One-leg balance time decreases ~2–5 seconds per decade

Low grip strength (<26 kg in men, <16 kg in women) predicts incident disability and mortality (hazard ratios 1.5–2.0 in meta-analyses)

**Cognitive function:**

Processing speed (DSST) declines ~0.5–1.0 correct symbols per year after age 40

Episodic memory (RAVLT) declines ~0.1–0.3 words recalled per year after age 50

Executive function (Trail Making B) increases ~2–5 seconds per decade (slower performance)

These cognitive declines accelerate after age 60

**Physiological function:**

FEV1 declines ~20–30 mL per year after age 30

Fasting glucose increases ~0.1–0.2 mmol/L per decade after age 40

HbA1c increases ~0.1–0.2% per decade after age 40

Systolic blood pressure increases ~5–10 mmHg per decade until age 70, then plateaus or declines

Bone mineral density declines ~0.5–1.0% per year after age 50 in women, ~0.3–0.5% per year in men

**Endocrine function:**

Morning cortisol increases slightly with age (~10–20% higher in 70-year-olds vs 30-year-olds)

Total testosterone declines ~1–2% per year in men after age 30

Oestradiol declines sharply at menopause (~50–80% drop over 2–5 years)

IGF-1 declines ~1–2% per year after age 30

**Immune function:**

CRP increases ~0.1–0.3 mg/L per decade after age 40

IL-6 increases ~0.1–0.2 pg/mL per decade after age 40

Higher CRP (>3 mg/L) and IL-6 (>2 pg/mL) predict mortality (hazard ratios 1.3–1.7)

In plain English, here is what the typical age-related decline looks like for a 50-year-old compared to a 70-year-old:

**Grip strength:** A 70-year-old man can grip about 10–20 kg less than a 50-year-old — roughly the weight of a large bag of groceries

**Gait speed:** A 70-year-old walks about 0.2–0.4 m/s slower — the difference between a brisk walk (1.2 m/s) and a slow stroll (0.8 m/s)

**Processing speed:** A 70-year-old completes about 10–20 fewer digit-symbol pairs in 90 seconds — roughly the difference between finishing a crossword in 10 minutes vs 15 minutes

**Lung function:** A 70-year-old has about 0.5–1.0 litres less FEV1 — enough to notice when climbing stairs

**Inflammation:** A 70-year-old has CRP levels about 1–3 mg/L higher — the difference between a low-risk (<1 mg/L) and moderate-risk (1–3 mg/L) category for cardiovascular disease

These are population averages. Individual trajectories vary widely — some 70-year-olds have biomarkers of a 50-year-old, and vice versa.

**Acknowledged by authors:**

The panel is a "proposed" set, not a validated composite — it needs testing in prospective studies

Some domains (e.g., dexterity, balance) have weaker evidence than others

The panel may need adaptation for different populations (e.g., very old, frail, non-Western)

Biomarkers were selected for feasibility in field settings, potentially missing more sensitive markers

The panel does not include emerging biomarkers (e.g., epigenetic clocks, telomere length) that were not well-established at the time

**Critical reader observations:**

**No formal consensus method:** The expert workshop lacked a Delphi process or quantified agreement, so the recommendations reflect the authors' synthesis plus informal expert input, not a rigorous consensus

**Publication date (2015):** The field has advanced significantly since 2015. Epigenetic clocks (e.g., Horvath clock, PhenoAge), DNA methylation markers, and multi-omics panels are now available and may outperform these traditional biomarkers. The panel is still useful but not state-of-the-art

**No composite score:** The authors recommend measuring all biomarkers but do not provide a formula to combine them into a single "ageing score." This limits practical use for self-experimenters who want a single number to track

**Population bias:** The evidence base is predominantly from Western, white, high-income populations. Normative values may not apply to other ethnicities or socioeconomic groups

**Sex differences:** The panel treats men and women similarly, but ageing trajectories differ markedly (e.g., menopause effects, testosterone decline). Separate norms are needed but not fully provided

**No longitudinal validation:** The panel was not tested in a new cohort to see if it predicts outcomes better than individual biomarkers

**Cost and accessibility:** Some tests (DXA scan for bone density, spirometry, blood biomarkers) require equipment or lab access, limiting self-experiment use

For someone running their own n=1 experiment to track healthy ageing:

### What to test

Measure the following core panel at baseline and at regular intervals (e.g., every 6–12 months). You can start with the most accessible tests and add others as resources allow:

**Tier 1 (low cost, home-based):**

Grip strength (hand dynamometer, ~$20–50)

Gait speed (timed 4-metre walk, stopwatch)

Chair rise time (timed 5 stands, stopwatch)

One-leg balance (timed 30 seconds, stopwatch)

Body weight and height (for BMI)

Resting heart rate and blood pressure (home monitor, ~$30–50)

Self-reported cognitive function (e.g., Cognitive Failures Questionnaire)

**Tier 2 (moderate cost, lab or clinic):**

Fasting glucose and HbA1c (blood test, ~$20–50)

CRP, IL-6 (blood test, ~$30–60)

Cortisol (morning saliva or blood, ~$30–50)

Testosterone (men) or oestradiol (women) (blood test, ~$50–100)

IGF-1 (blood test, ~$50–100)

Spirometry (FEV1, FVC) (clinic visit, ~$50–100)

Bone mineral density (DXA scan, ~$100–200)

**Tier 3 (higher cost, comprehensive):**

Full cognitive battery (e.g., NIH Toolbox, ~$200–500 for online version)

Dexterity (Purdue Pegboard, ~$100–200 for the board)

### Minimum meaningful duration

**For tracking age-related decline:** Measure every 12 months for at least 3–5 years to see trends above measurement noise

**For testing an intervention (e.g., exercise, diet, supplement):** Measure at baseline, then after 3–6 months of intervention. Some biomarkers (CRP, glucose, grip strength) can change in 3 months; others (bone density, cognitive function) may need 6–12 months

**For a single time point:** Useless — you need at least two measurements to calculate rate of change

### What to measure (specific metrics)

Track these as your primary outcomes:

| Domain | Primary metric | Target direction |

|--------|----------------|------------------|

| Strength | Grip strength (kg, best of 2 trials per hand) | Higher = better |

| Mobility | Gait speed (m/s, usual pace over 4m) | Higher = better |

| Balance | One-leg stand (seconds, max 30) | Higher = better |

| Processing speed | Digit Symbol Substitution Test (correct in 90s) | Higher = better |

| Memory | Rey Auditory Verbal Learning Test (words recalled) | Higher = better |

| Lung function | FEV1 (litres) | Higher = better |

| Glucose metabolism | HbA1c (%) | Lower = better |

| Inflammation | CRP (mg/L) | Lower = better |

| Endocrine | Morning cortisol (nmol/L) | Mid-range (~200–500 nmol/L) |

| Endocrine | Testosterone (men, nmol/L) | Higher = better (within range) |

### Key confounds to control for

**Time of day:** Measure biomarkers at the same time of day (morning for cortisol, fasting for glucose, afternoon for grip strength)

**Recent activity:** No vigorous exercise 24 hours before blood tests (raises CRP, cortisol)

**Illness:** Delay testing if you have a cold, infection, or injury (raises CRP