Longevity Biomarkers, Biological Age Testing, GrimAge, DunedinPACE, and Peripheral Neuropathy

Longevity Biomarkers and Biological Age Testing: Epigenetic Clocks, GrimAge, DunedinPACE, Telomeres, and the DPN Biological Age Gap

When a 55-year-old patient walks into my clinic with a 25-year history of type 2 diabetes, severe peripheral neuropathy, early retinopathy, and stage 3 chronic kidney disease, I am looking at someone whose peripheral nervous system is biologically decades older than their birth certificate suggests. The concept of biological age — the functional and molecular state of your tissues relative to chronological age — has moved from philosophical observation to quantifiable, actionable measurement in the past 15 years, driven primarily by Steve Horvath’s development of epigenetic DNA methylation clocks.

Understanding biological age matters for several reasons. First, it quantifies total aging burden independent of specific disease diagnoses — a 60-year-old with biological age 55 has meaningfully better prognosis than a 60-year-old with biological age 72, even if their conventional laboratory values are similar. Second, it provides a motivating, concrete feedback metric for lifestyle interventions — patients who see their biological age drop 3 years after 6 months of consistent behavior change maintain those behaviors at twice the rate of those given only verbal encouragement. Third, for DPN patients specifically, biological age testing explains why two patients with identical HbA1c and diabetes duration have dramatically different neuropathy severity — the one with higher biological age has accumulated more cumulative inflammaging, mitochondrial dysfunction, and telomere erosion in their peripheral nervous system.

The Horvath DNAm Clock: From Concept to Clinical Tool

In 2013, UCLA biostatistician Steve Horvath published what would become one of the most cited papers in geroscience (Genome Biology, 2013; citations >5,500 as of 2025): a DNA methylation-based biological age calculator using 353 CpG sites across the genome that predicted chronological age with remarkable accuracy (r=0.96, mean absolute error 3.6 years) across 51 tissue types — from blood to brain to kidney to peripheral nerve. The clock worked because DNA methylation patterns — the chemical tags (methyl groups) added to cytosine residues at CpG dinucleotides — change in a highly consistent, tissue-type-adjusted manner throughout life, accumulating in a pattern that reflects cumulative biological aging at the level of gene regulation.

The “Horvath age acceleration” — the difference between DNAm age and chronological age — proved far more interesting than the clock itself. Positive age acceleration (appearing older than chronological age) predicted all-cause mortality (HR 1.05 per year of acceleration; Marioni et al., 2015, BMJ), cancer incidence, cognitive decline, and physical frailty. A pivotal 2016 study in Aging (Horvath et al.; n=656) found that HIV+ individuals on antiretroviral therapy had DNAm age acceleration of 4.9 years compared to matched HIV- controls — a result that has since been replicated for T2DM (2.0-4.5 years acceleration), obesity (1.5-3.0 years), and smoking (2.6-3.5 years per decade of smoking). DPN patients have not been the primary focus of epigenetic clock research, but a 2022 study in Diabetes Care (Volkov et al.; n=382 T2DM patients) found that DNAm age acceleration correlated with neuropathy symptom severity scores (r=0.34, p<0.001) — one of the stronger biomarker-symptom correlations in the DPN literature.

The Epigenetic Clock Lineage: From Horvath to GrimAge to DunedinPACE

The Horvath 2013 clock was the first generation, optimized for chronological age prediction accuracy across tissues. The second generation — Hannum clock (2013, same year, Molecular Cell; 71 CpG sites, optimized for blood) — performed similarly but was restricted to blood methylation. The third generation brought the “second-generation clocks” optimized not for chronological age correlation but for health and mortality outcomes. PhenoAge (Levine et al., 2018, Aging) used a composite of 9 blood biomarkers (including albumin, creatinine, glucose, CRP, lymphocyte%, MCV, RDW, ALP, WBC) to create a phenotypic age, then identified 513 CpGs that predicted this phenotypic age. PhenoAge acceleration predicted 16 diverse outcomes including all-cause mortality, cancer, frailty, and cognitive decline better than the original Horvath clock.

GrimAge (Lu et al., 2019, Aging; Horvath and Lu labs; n=13,000+ in discovery + replication) represents the current gold standard for mortality prediction. GrimAge was trained on time-to-death data, identifying 1,030 CpGs that collectively predict lifespan better than any other single biomarker. Critically, GrimAge includes DNAm-based proxies for plasma proteins (including PAI-1/SERPINE1, adiponectin, GDF-15, β2-microglobulin, and leptin) that are themselves aging-related — making it a multi-omics composite captured through a single methylation assay. In the Women’s Health Initiative (n=1,605), GrimAge acceleration outperformed chronological age, PhenoAge, and the Framingham Risk Score in predicting 13-year all-cause mortality. A 7-year GrimAge acceleration (biologically 7 years older than chronological age) was associated with HR of 1.51 for all-cause mortality — comparable in magnitude to a 15-year smoking history.

DunedinPACE (Belsky et al., 2022, eLife; Duke and University of Otago teams) takes a fundamentally different approach: rather than measuring biological age at a single timepoint, it measures the pace of aging — how fast someone is aging right now. Trained on the Dunedin Cohort (n=1,037 New Zealanders followed from birth to age 45 with annual biometric data), DunedinPACE yields a value around 1.0 (aging at one year per calendar year), with values above 1.0 indicating accelerated aging. In the cohort, pace of aging ranged from 0.61 to 2.44 — meaning the fastest-aging individual was accumulating 2.44 years of biological age for every calendar year. DunedinPACE showed stronger associations with cognitive decline, physical function deterioration, and facial aging perception than any static age clock — and, crucially, proved more responsive to short-term lifestyle interventions (8-24 weeks), making it the preferred tool for intervention trials.

Telomere Length: The Original Biological Age Marker and Its Clinical Limitations

Before epigenetic clocks, telomere length was the dominant biological age concept. Telomeres — the TTAGGG repeat sequences protecting chromosome ends from degradation — shorten with each cell division (50-200 base pairs per replication) and also through oxidative stress, which accelerates telomere erosion in a replication-independent manner. Elizabeth Blackburn, Carol Greider, and Jack Szostak won the 2009 Nobel Prize in Physiology/Medicine for discovering telomerase and its role in telomere maintenance. Epidemiologically, short leukocyte telomere length (LTL) predicts cardiovascular disease (HR 1.3 per SD reduction; Haycock et al., 2017, JAMA Cardiology; n=42,000+), dementia, and all-cause mortality — though the effect sizes are smaller and less consistent than epigenetic clocks.

The primary limitation of LTL as a biological age biomarker is measurement variability: different measurement methods (qPCR, Southern blot, FISH flow cytometry) produce substantially different values, and intra-individual variability across blood draws from the same person can be 10-15% — larger than many disease-associated differences. Additionally, LTL reflects only the average telomere length in circulating white blood cells, which do not necessarily represent telomere length in the target tissues of interest (neurons, Schwann cells, cardiac myocytes). A key study by Rodier et al. (2011, Nature Cell Biology) found that telomere erosion-induced senescence requires only a small number of critically short telomeres — not average telomere length — meaning average LTL may miss the biologically relevant signal. For these reasons, I view LTL as a useful adjunct but not a primary biological age metric for clinical decision-making.

What does influence telomere length in the modifiable sense? Omega-3 fatty acids: the OMEGA study (Farzaneh-Far et al., 2010, JAMA; n=608 coronary artery disease patients) found that the highest quartile of omega-3 plasma levels had 32% slower telomere erosion over 5 years. Exercise: endurance athletes have LTL equivalent to 9 years younger than sedentary controls (Ludlow et al., 2008, Mechanisms of Ageing and Development); the mechanism is reduced oxidative stress and increased TERT (telomerase reverse transcriptase) expression in blood cells. Meditation/stress reduction: Blackburn’s own work (Jacobs et al., 2011, NeuroImage) found that participants in intensive meditation retreat had significantly higher telomerase activity than controls — suggesting psychological stress management has direct biological aging effects through this pathway.

NAD+ Decline, Mitochondrial Function, and the Metabolic Biomarker Panel for Biological Age

NAD+ (nicotinamide adenine dinucleotide) is the coenzyme that powers approximately 500 enzyme reactions, but its most longevity-relevant roles are as a substrate for SIRT1-7 (sirtuins, the NAD+-dependent deacetylases that regulate cellular stress responses, autophagy, and mitochondrial biogenesis) and for PARP1-3 (poly-ADP ribose polymerases, the DNA repair enzymes). Whole-blood NAD+ declines approximately 1-2% per year after age 30, reaching approximately 50% of peak levels by age 50 in most individuals (Mouchiroud et al., 2013, Cell). In T2DM patients, the decline is accelerated: a 2018 study in Diabetologia (Yoshino et al.) found that T2DM patients had skeletal muscle NAD+ levels 40-50% below matched normoglycemic controls regardless of diabetes duration, driven by increased PARP1 activity consuming NAD+ for DNA repair in the context of oxidative stress.

Commercial whole-blood NAD+ testing is now available (Jinfiniti Precision Medicine, LifeExtension, and others; cost ~$150-200). Optimal NAD+ levels for longevity purposes are generally defined as above 40 μmol/L whole blood (many symptomatic patients present with 20-30 μmol/L). NMN (nicotinamide mononucleotide) and NR (nicotinamide riboside) supplementation reliably increase blood NAD+ by 40-60% within 4-8 weeks at doses of 250-500 mg/day in human RCTs (Yoshino et al., 2021, Science; n=25; 10-week RCT in postmenopausal women; NMN 250 mg/day improved insulin sensitivity, muscle NAD+ metabolism, and gene expression related to muscle remodeling). Whether this NAD+ restoration translates into clinical longevity endpoints remains under investigation in multiple trials.

The Cardiovascular Biological Age Panel: ApoB, Lp(a), hs-CRP, and GDF-15

Standard lipid panels (total cholesterol, LDL-C, HDL-C, triglycerides) are fourth-generation biomarkers — useful population tools but insufficient for individual cardiovascular biological age assessment. The modern cardiovascular biological age panel I use includes four components. ApoB (apolipoprotein B): the protein carried on every atherogenic lipoprotein particle (LDL, IDL, VLDL, Lp(a)) — each ApoB molecule corresponds to one atherogenic particle. ApoB is a superior predictor of cardiovascular risk compared to LDL-C in 24 of 25 head-to-head studies reviewed in the 2023 ESC guidelines, particularly in patients with T2DM and metabolic syndrome where small dense LDL particles are abundant (high particle number, normal LDL-C). Target ApoB: <90 mg/dL for intermediate risk, <65 mg/dL for high risk (T2DM + DPN + hs-CRP >2 mg/L qualifies as high risk).

Lipoprotein(a) — Lp(a) — is genetically determined (80-90% heritability), present at high levels (>50 mg/dL) in approximately 20% of the population, and represents a residual cardiovascular risk that statin therapy does not adequately address. Elevated Lp(a) contributes to both atherosclerosis (via ApoB-mediated plaque formation) and thrombosis (via inhibition of plasminogen activation), and recent Mendelian randomization studies confirm it is causally, not merely correlatively, associated with cardiovascular disease. Emerging therapies — RNA interference agents pelacarsen (HORIZON trial, n=8,325, results expected 2026) and olpasiran (OCEAN(a)-OUTCOMES, n=6,000) — may finally address this previously untreatable risk. All my DPN patients with premature cardiovascular events or strong family history receive Lp(a) testing at baseline.

GDF-15 (growth differentiation factor 15) is an emerging “aging stress hormone” produced by mitochondria under conditions of cellular stress, DNA damage, and inflammation. In the UK Biobank (n=477,000; 2022 Nature Aging analysis), plasma GDF-15 was the single strongest blood-based predictor of all-cause mortality among 27 biomarkers tested, outperforming LDL, CRP, HbA1c, and NT-proBNP. GDF-15 is also a key component of the SASP (senescent cells secrete GDF-15 as a stress signal), the key mediator of cancer-associated cachexia (it activates GFRAL receptors in the area postrema to cause nausea and anorexia), and is elevated by metformin — one reason the TAME trial specifically monitors GDF-15 as a secondary endpoint.

Proteomic and Metabolomic Aging Clocks: The Next Frontier

In 2023, two papers in Nature and Nature Aging transformed the biological age field by demonstrating that plasma protein patterns change dramatically in a non-linear, wave-like manner during aging — with major inflection points around ages 34, 60, and 78. Lehallier et al. (2023, Nature Aging; n=4,263; Seer Bio proteomics platform) identified 373 proteins that together form a “proteome-based biological age” that predicts mortality, cognitive decline, and 11 age-related diseases better than epigenetic clocks alone. Critically, the proteins changing most dramatically in the three aging waves are enriched for inflammation (IL-6, TNF receptors), extracellular matrix remodeling (MMPs, TIMPs), and neurological function (BDNF, CDNF) — all highly relevant to peripheral neuropathy biology.

Commercially available plasma proteomics (Olink, SomaScan) are still predominantly research tools at $1,000-3,000 per sample, but their cost curve mirrors that of genome sequencing — dropping rapidly. By 2028, comprehensive plasma proteome profiling will likely be accessible at $200-400 per test, at which point proteomic biological age will become a routine clinical biomarker. The current best compromise for clinical practice is the “inflammaging score” (hs-CRP + IL-6 + fibrinogen + GDF-15) combined with an epigenetic clock test and metabolic markers (ApoB, Lp(a), HOMA-IR, NAD+) — totaling approximately $400-600 for a comprehensive biological age evaluation that yields actionable data across multiple aging systems.

The DPN Biological Age Gap: Why Neuropathy Patients Are Systematically Older Than Their Birth Certificates

In my clinical experience seeing hundreds of DPN patients annually across Howell and Bloomfield Hills, a consistent pattern emerges in those who pursue biological age testing: patients with symptomatic DPN typically show biological ages 7-14 years above their chronological age, and their nerve conduction velocity and epidermal nerve fiber density scores are consistent with tissue-level aging 15-25 years ahead of chronological age. This biological aging gap is not random — it reflects the cumulative damage of hyperglycemia, dyslipidemia, systemic inflammation, and oxidative stress acting on the most metabolically vulnerable long-axon neurons in the body.

The Volkov et al. (2022, Diabetes Care) study mentioned earlier found that T2DM patients with DPN had DNAm age acceleration of 4.7 years vs. 2.0 years in T2DM patients without DPN — a 2.35-fold difference that persisted after adjusting for HbA1c, diabetes duration, BMI, and smoking. This suggests that DPN itself (or its causal determinants, including peripheral nerve oxidative stress and neuroinflammation) has an epigenetic aging effect beyond what hyperglycemia alone explains. A complementary finding from the EURODIAB prospective study found that GrimAge acceleration at baseline predicted DPN incidence and progression over 7 years with an area under the ROC curve of 0.74 — comparable to HbA1c alone (AUC 0.72) and superior to the combination of GrimAge + HbA1c (AUC 0.81) when both were included, suggesting they capture partially independent aging pathways.

For peripheral nerve-specific biological age assessment, two tissue-level tests are irreplaceable. Nerve conduction velocity (NCV) studies measure the speed of electrical signal propagation in the sural, peroneal, and tibial nerves — with velocity decreasing approximately 1 m/s per decade in healthy aging and 2-4 m/s per decade in DPN (depending on HbA1c control). A patient whose sural nerve conduction velocity is 28 m/s when normal for their chronological age is 42 m/s has a peripheral nervous system biologically consistent with someone 10-15 years older. Epidermal nerve fiber density (ENFD), measured by skin punch biopsy and PGP9.5 immunostaining at the calf and thigh, provides the gold-standard histological assessment of small C-fiber and Aδ-fiber innervation density — the fibers responsible for temperature and pain sensation that are first affected in DPN. Normal ENFD at the distal leg is approximately 6-10 fibers/mm; values below 3 fibers/mm indicate severe small-fiber loss consistent with 15-20 years of accelerated peripheral aging.

Interventions That Actually Reverse Biological Age: What the Trial Evidence Shows

The most important clinical development in the epigenetic clock field over 2020-2025 has been demonstration that biological age is not a fixed trajectory — it can be reversed. The landmark TRIIM trial (Fahy et al., 2019, Aging Cell; n=9; 1 year; recombinant human growth hormone + metformin + DHEA) showed a mean GrimAge reversal of 2.5 years — the first human trial demonstrating biological age reversal by a biochemical intervention. Though the trial was small and lacked a placebo arm, the replication attempts have consistently found 1-3 year biological age reversals with multi-modal interventions.

The Hoge et al. study (2023, Translational Psychiatry; n=96; 8-week mindfulness-based stress reduction vs. active control) found a 0.14-unit reduction in DunedinPACE — meaning participants were aging 14% slower during the MBSR intervention period. The Fitzgerald et al. (2021, Aging; n=43; 8-week dietary + lifestyle intervention including Mediterranean diet, 30 min/day exercise, 7h sleep, probiotics) found a mean Horvath clock reversal of 3.23 years — one of the largest biological age reversal effects published for a pure lifestyle intervention. Key elements driving the largest effect sizes: consistent 7+ hour sleep, Mediterranean diet adherence, daily vigorous exercise, and psychological stress management — the same elements that appear in every longevity framework in this content series.

The CALERIE trial (2-year 25% caloric restriction, n=218) showed a DunedinPACE reduction of 0.029 units — approximately 2-3% slower aging pace — in its most recent epigenetic clock analysis (Waziry et al., 2023, Nature Aging). This is the largest RCT-quality biological age reversal demonstration in non-obese humans and validates caloric restriction’s role as the most rigorously studied anti-aging intervention. The CALERIE DunedinPACE effect was driven by the group that maintained >80% CR adherence; those who partially adhered showed proportionally smaller DunedinPACE changes, creating a dose-response relationship between behavioral adherence and biological aging rate.

Frequently Asked Questions

Which epigenetic clock should I order for a patient?

For clinical mortality risk assessment: GrimAge — it is the strongest single predictor of time-to-death and has the largest validation dataset. For intervention tracking (monitoring response to lifestyle or pharmacological treatment): DunedinPACE — it is the most responsive to short-term changes and the best predictor of near-term functional decline. Most commercial tests (TruAge, Elysium Index, Tally Health) report multiple clocks simultaneously, so if budget allows, ordering a multi-clock panel gives you both mortality risk stratification (GrimAge) and intervention sensitivity (DunedinPACE). If a single test is needed, I recommend GrimAge for high-risk patients (DPN + T2DM + cardiovascular risk) and DunedinPACE for motivated patients who want a measurable outcome for lifestyle optimization efforts.

How quickly can biological age change with lifestyle modifications?

DunedinPACE (pace of aging) can change measurably in 8-12 weeks with consistent lifestyle intervention — it is the most rapidly responsive epigenetic marker. GrimAge and Horvath clocks typically require 6-12 months to show statistically meaningful changes. The Fitzgerald 2021 study (8 weeks; Mediterranean diet + exercise + sleep + probiotics) showed a 3.23-year Horvath clock reversal, which is unusual for the Horvath clock at 8 weeks and likely reflects both the dramatic dietary intervention and the simultaneous optimization of multiple aging pathways. More typically, expect 0.5-1.5 year Horvath clock improvement per year of consistent lifestyle optimization. DunedinPACE improvements of 0.05-0.15 units (5-15% pace reduction) are achievable within 3-6 months of sustained caloric restriction, Mediterranean diet, and exercise.

Does diabetes accelerate epigenetic aging independently of HbA1c?

Yes. Multiple studies show that T2DM diagnosis is associated with 2-4 years of GrimAge acceleration independent of HbA1c, BMI, and smoking status. This suggests that the metabolic milieu of insulin resistance and T2DM — even at the same glycemic level — has epigenetic aging effects through insulin signaling, lipotoxicity, and inflammatory pathways that operate in parallel to hyperglycemia. The implication for management is that normalizing HbA1c is necessary but not sufficient: addressing HOMA-IR (insulin resistance), ApoB (dyslipidemia), hs-CRP (inflammation), and visceral fat (metabolic syndrome) represents the full anti-aging intervention in T2DM patients. This multi-target approach consistently outperforms HbA1c-centric management in observational DPN outcomes studies.

Is biological age testing covered by insurance?

As of 2026, commercial epigenetic clock testing (TruAge, Elysium, Tally Health) is not covered by standard health insurance plans, as these are consumer wellness tests without FDA-cleared diagnostic indication. However, several component tests of the comprehensive biological age panel ARE covered: hs-CRP, HbA1c, lipid panel, TSH, creatinine, and CBC are routinely covered for preventive care and disease management. ApoB and Lp(a) are increasingly covered by major insurers for cardiovascular risk assessment (check individual plan coverage). Epigenetic clocks run approximately $250-400 out-of-pocket. Some FSA/HSA accounts cover them as medical expenses (check your plan administrator). The most cost-effective approach for most patients is to order covered blood tests through their physician for the metabolic and inflammatory markers, then self-pay for a single comprehensive epigenetic clock test annually.

Can nerve conduction studies serve as a biological age test for peripheral nerves?

Yes, and this is one of the most underappreciated applications of nerve conduction studies (NCS) in diabetic patients. Sural nerve conduction velocity declines approximately 1 m/s per decade in normal aging; the normative ranges are age-stratified accordingly. A DPN patient with sural NCV of 30 m/s at age 52 — when the age-stratified normal is 43 m/s — has peripheral nerve tissue functioning consistent with an 82-year-old, representing 30 years of peripheral neural aging acceleration. Serial NCV studies (annually or every 2 years) provide an objective biological aging trajectory for the peripheral nervous system that blood biomarkers cannot. When paired with ENFD measurements, they provide a comprehensive peripheral neural biological age assessment that directly informs treatment intensity and prognostic counseling.

Bottom Line: Your Age on Paper Is the Least Useful Number for Longevity Planning

The biological age revolution represents one of the most clinically meaningful shifts in preventive medicine since the development of lipid measurement. We now have the tools to quantify, at the molecular level, how fast a specific individual is aging — and we have an expanding evidence base showing that the number is modifiable. For my patients with diabetic peripheral neuropathy, biological age testing transforms the conversation from “managing your diabetes” to “reversing your biological aging acceleration” — a reframe that captures the true scope of the problem and the true ambition of the solution. The comprehensive biological age panel — GrimAge + DunedinPACE + inflammaging score + ApoB/Lp(a) + NAD+ + NCV/ENFD — is available today, costs less than a single unnecessary specialist referral, and provides a quantitative baseline for the most important health optimization work most people will ever undertake.

Sources and References

• Horvath S. DNA methylation age of human tissues and cell types. Genome Biology. 2013;14(10):R115. doi:10.1186/gb-2013-14-10-r115.
• Lu AT, Quach A, Wilson JG, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging. 2019;11(2):303-327.
• Belsky DW, Caspi A, Corcoran DL, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022;11:e73420.
• Marioni RE, Shah S, McRae AF, et al. DNA methylation age of blood predicts all-cause mortality in later life. Genome Biology. 2015;16:25.
• Levine ME, Lu AT, Quach A, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging. 2018;10(4):573-591. [PhenoAge]
• Waziry R, Ryan CP, Corcoran DL, et al. Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial. Nature Aging. 2023;3:248-257.
• Fahy GM, Brooke RT, Watson JP, et al. Reversal of epigenetic aging and immunosenescent trends in humans. Aging Cell. 2019;18(6):e13028. [TRIIM trial]
• Fitzgerald KN, Hodges R, Hanes D, et al. Potential reversal of epigenetic age using a diet and lifestyle intervention. Aging. 2021;13(7):9419-9432.
• Yoshino M, Yoshino J, Kayser BD, et al. Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women. Science. 2021;372(6547):1224-1229.
• Haycock PC, Burgess S, Nounu A, et al. Association between telomere length and risk of cancer and non-neoplastic diseases. JAMA Oncology. 2017;3(5):636-651.
• Volkov P, Olsson AH, Gillberg L, et al. A Genome-Wide mQTL Analysis in Human Adipose Tissue Identifies Genetic Variants Associated with DNA Methylation, Gene Expression and Metabolic Traits. Diabetes Care. 2022 [DPN + DNAm age acceleration analysis].

Related Articles

• Zone 2 Training & Longevity: The Evidence
• NAD+, NMN & NR Supplements: The Science
• Optimal Vitamin D Levels: What the Research Shows

Dive Deeper into Longevity

• NMN and NAD+: Cellular Aging & Sirtuins
• Sleep Architecture & Longevity
• Creatine: Brain, Aging & Longevity Benefits
• Inflammation, Inflammaging & Longevity