Epigenetic
Epigenetic / DNA Methylation (DNAm)
Measure methylation patterns at CpG sites across the genome. The most mature and widely validated class of biological age clocks.
What it measures
DNA methylation beta values at specific CpG sites. First-generation clocks (Horvath, Hannum) predict chronological age; second/third-generation clocks (PhenoAge, GrimAge, DunedinPACE) predict mortality, morbidity, and pace of aging.
Strengths
- +Largest independent validation literature
- +Standardized arrays (Illumina 450k/EPIC)
- +Newer clocks (GrimAge, DunedinPACE) track health outcomes better than chronAge clocks
- +DunedinPACE designed as a pace-of-aging speedometer
Limitations
- −Cell-mixture confounding: saliva/cheek contain mixed immune/epithelial cells
- −Platform shift: new EPIC arrays may omit probes used by older clocks
- −A single number hides organ-specific heterogeneity
- −Point estimates lack per-person uncertainty quantification
Samples:BloodSalivaCheek swabDried blood spot
Best for: Longitudinal aging tracking, mortality/morbidity risk stratification, intervention studies requiring validated endpoints
Glycomic
Glycomic (IgG Glycosylation)
Analyze patterns of IgG N-glycans as markers of inflammaging. Provide a focused window into immune aging and inflammatory status.
What it measures
IgG glycosylation patterns (typically 20-30 N-glycan structures) from blood samples. Glycan profiles correlate with both chronological and biological aging-related physiology.
Strengths
- +Focused window into inflammaging
- +Evidence correlates with both chronological and biological age
- +Growing methodological work on analytical precision
- +Can be measured from simple finger-prick samples
Limitations
- −Narrower biomarker layer than multi-omic approaches
- −Lab-specific calibration required
- −Fewer independent validation studies than epigenetic clocks
- −Primarily reflects immune/inflammatory aging dimension
Samples:Finger-prick dried bloodVenous blood
Best for: Monitoring inflammatory aging trajectory, lifestyle intervention response tracking, complementary to epigenetic clocks
Phenotypic
Phenotypic / Physiological
Use routine clinical biomarkers, blood panels, and anthropometrics. Inexpensive, interpretable, and directly actionable.
What it measures
Standard clinical markers: CBC/CMP panels, lipids, inflammation markers (CRP, IL-6), anthropometrics, and functional tests. Some include grip strength, gait speed, and other physical measures.
Strengths
- +Inexpensive compared to omics-based clocks
- +Highly interpretable: you can directly address high glucose, inflammation, lipids
- +Clinically actionable: maps to standard medical interventions
- +Low-cost risk proxy with clear what moved your score feedback
Limitations
- −May capture disease burden rather than pure aging
- −Less precise at measuring aging-specific biology
- −Non-black-box but limited to clinical markers
- −State availability restrictions for some US products
Samples:Venous blood drawClinical assessment
Best for: Cost-conscious consumers wanting actionable health feedback, general wellness monitoring, near-term risk reduction
Transcriptomic
Transcriptomic
Use gene expression profiles to predict age or aging phenotypes. Sensitive to immune and metabolic state but also to acute perturbations.
What it measures
Gene expression levels via bulk RNA-seq, microarrays, or single-cell approaches. Transcriptomic clocks capture immune/metabolic state and can infer aging trajectories from blood or stool samples.
Strengths
- +Sensitive to immune and metabolic state changes
- +Research-grade age and health inference
- +Can capture acute physiological responses
- +Large-scale models exist in research literature
Limitations
- −Sensitive to acute physiology (infection, stress) which can be noise
- −Cell composition and technical pipelines are major confounders
- −Commercially appears mostly inside broader multi-omic products
- −Cross-platform portability is limited
Samples:BloodStoolTissue biopsy
Best for: Research-grade aging studies, multi-omic panels, immune/metabolic state monitoring
Proteomic
Proteomic
Measure plasma protein panels to predict morbidity, mortality, and organ-level aging signatures.
What it measures
Panels of plasma proteins measured by aptamer-based (SomaScan) or antibody-based (Olink) platforms. Proteomic age signatures can predict disease risk, mortality, and organ-specific aging.
Strengths
- +Strong links to disease and mortality (closer to physiology)
- +Large-biobank proteomic clocks demonstrate morbidity/mortality prediction
- +Can provide organ-level and pathway-level signatures
- +Demonstrated incremental value beyond chronological age
Limitations
- −Platform dependence (aptamer vs antibody panels)
- −Cross-platform portability is limited
- −Cost remains a barrier for consumer use
- −Primarily available through research/clinical partnerships
Samples:Venous blood
Best for: Mortality/morbidity risk stratification, organ-level aging assessment, clinical research
Microbiome
Microbiome
Use gut and oral microbial profiles to estimate biological age. Responsive to diet and lifestyle but highly variable across populations.
What it measures
Microbial taxonomic and functional profiles via 16S sequencing, metagenomics, or metatranscriptomics. Models can predict age reasonably well but are strongly influenced by geography, diet, and antibiotics.
Strengths
- +Some age prediction capability
- +Exploratory links to diet and disease
- +Responsive to lifestyle interventions
- +Non-invasive sample collection
Limitations
- −Highly confounded by geography, diet, and antibiotics
- −Portability across populations is challenging
- −Many microbiome tests offer age-like insights without rigorous clock validation
- −Short-term fluctuations can be significant
Samples:StoolOral swab
Best for: Exploratory gut health monitoring, diet/lifestyle response tracking, research applications
Multi-omic
Composite / Multi-Omic
Combine multiple biomarker layers (DNA methylation, proteins, metabolites, clinical markers) for potentially broader signal coverage.
What it measures
Combined omics data plus clinical markers. Aims to capture multiple dimensions of aging simultaneously, though this comes with opacity and dataset shift risks.
Strengths
- +Potentially better signal coverage across aging dimensions
- +Can combine strengths of individual clock types
- +May capture aging heterogeneity better than single-layer clocks
Limitations
- −Increased opacity: harder to interpret what drives the score
- −Dataset shift risk when combining platforms
- −Many proprietary composites lack independent benchmarking
- −Higher cost due to multiple assays
Samples:BloodMultiple sample types
Best for: Comprehensive aging assessment, research settings with multiple data streams
Immune
Immune / Inflammatory
Measure immune cell subsets, cytokines, and immunome profiles to define immune age or inflammatory age.
What it measures
Immune cell subset composition, cytokines/chemokines, or multi-omic immunome profiles. Derived from frameworks like the Stanford 1,000 Immunomes project.
Strengths
- +Predicts immune health, mortality, and frailty
- +Immune age trajectories (IMM-AGE) predict all-cause mortality beyond standard risk factors
- +Grounded in established immunology research
- +Directly relevant to inflammaging
Limitations
- −Primarily available through clinic/partner models, not direct-to-consumer
- −Pricing often not publicly listed
- −Fewer consumer-facing validation studies
- −Requires standard blood draw
Samples:Venous blood
Best for: Immune health monitoring, inflammation tracking, clinical aging research
Telomere
Telomere-Based
Measure leukocyte telomere length as a proxy for cellular aging. Correlates with age but has high measurement variability.
What it measures
Leukocyte telomere length via qPCR, flow-FISH, or TRF methods. Telomere length correlates with age and some health outcomes but many experts consider it a weak or context-specific aging proxy.
Strengths
- +Addresses specific telomere biology questions
- +Long research history in aging biology
- +Can complement other clock types as auxiliary data
Limitations
- −High measurement variability across methods
- −Tissue specificity limits interpretation
- −Many experts consider it a weak standalone aging clock
- −Weaker than multi-biomarker models for general aging prediction
Samples:Blood
Best for: Telomere biology research, auxiliary aging data alongside other clocks
Imaging
Imaging-Based
Use deep learning on facial photos, chest X-rays, brain MRI, and other imaging to estimate biological age.
What it measures
Deep learning features extracted from medical images. FaceAge (facial photos), CXR-Age (chest radiographs), and brain MRI models have shown prognostic value in specific clinical cohorts.
Strengths
- +Prognostic signals in specific domains (oncology, cardiopulmonary)
- +Some models validated in large clinical cohorts
- +Non-invasive (especially facial photo approaches)
Limitations
- −Privacy and ethics concerns, especially for facial photo approaches
- −Requires careful calibration and bias auditing
- −Not clearly sold as standalone consumer tests
- −Domain-specific rather than general aging clocks
Samples:Facial photoChest X-rayBrain MRI
Best for: Clinical research, domain-specific prognostic assessment, not yet consumer-ready