Epigenetic
Epigenetic / DNA Methylation (DNAm)
Measure chemical tags on your DNA that change as you age — like wear marks on a biological clock. The most mature and widely validated class of aging clocks.
What it measures
Small chemical modifications on your DNA that accumulate over time. Early clocks estimated your calendar age from these patterns; newer clocks go further, predicting disease risk, lifespan, and how fast you are aging right now.
Strengths
- +Most extensively studied and independently validated
- +Uses standardized lab technology across providers
- +Newer clocks predict health outcomes, not just calendar age
- +Some clocks act as a speedometer — measuring your pace of aging
Limitations
- −Saliva and cheek samples mix different cell types, which can skew results
- −Newer lab equipment may not be fully compatible with older clock models
- −A single number can hide differences in how individual organs are aging
- −Results don't include a personal margin of error
Samples:BloodSalivaCheek swabDried blood spot
Best for: Tracking aging over time, understanding disease and mortality risk, measuring whether lifestyle changes are making a difference
Technical details›
Measures DNA methylation beta values at specific CpG sites. First-generation clocks (Horvath: pan-tissue, 353 CpGs; Hannum: blood, 71 CpGs) predict chronological age. Second/third-generation clocks (PhenoAge, GrimAge, DunedinPACE) are trained on mortality, morbidity, and pace-of-aging outcomes. All models use elastic net / penalized regression. Measured on standardized Illumina 450k/EPIC arrays. Platform shift risk: newer EPIC v2 arrays may omit probes used by older clock models. Cell-mixture confounding is a known issue — saliva and cheek swab samples contain mixed immune and epithelial cells.
Common models: Horvath, Hannum, PhenoAge, GrimAge, DunedinPACE
Glycomic
Glycomic (IgG Glycosylation)
Analyze sugar molecules attached to immune proteins in your blood. These patterns shift as you age and reflect how much low-grade inflammation your body carries.
What it measures
The sugar coating on your antibodies (immune proteins). The pattern of these sugars changes with age and correlates with chronic inflammation — a key driver of age-related disease.
Strengths
- +Provides a focused window into immune aging and inflammation
- +Correlates with both calendar age and biological aging
- +Growing body of research on measurement precision
- +Can be measured from a simple finger-prick at home
Limitations
- −Captures only one dimension of aging (immune/inflammatory)
- −Each lab needs its own calibration, limiting comparability
- −Fewer independent studies than epigenetic clocks
- −Doesn't reflect aging in organs beyond the immune system
Samples:Finger-prick dried bloodVenous blood
Best for: Tracking inflammatory aging over time, measuring response to lifestyle changes, complementing other clock types
Technical details›
Measures IgG N-glycosylation patterns — typically 20–30 glycan structures — from blood samples via HPLC or mass spectrometry. Glycan profiles (e.g. ratio of agalactosylated to digalactosylated species) correlate with both chronological and biological age. Models use regression/score composites across ~27 IgG glycan structures with lab-specific calibration.
Common models: GlycanAge and related composite scores
Phenotypic
Phenotypic / Physiological
Use standard blood tests and physical measurements you may already get at a check-up. The most affordable and easiest-to-understand approach.
What it measures
Familiar health markers — blood sugar, cholesterol, inflammation levels, blood cell counts — and sometimes physical tests like grip strength or walking speed. These are combined into a single age estimate.
Strengths
- +Much cheaper than molecular-level clocks
- +Easy to understand: you can see exactly which markers are off
- +Directly actionable — results map to things your doctor already treats
- +Clear feedback on what moved your score
Limitations
- −May reflect current disease more than underlying aging
- −Less precise at measuring aging-specific biology
- −Limited to what standard blood tests can reveal
- −Some US products have state availability restrictions
Samples:Venous blood drawClinical assessment
Best for: People who want affordable, actionable health feedback and general wellness monitoring
Technical details›
Uses standard clinical markers: CBC/CMP panels, lipids, inflammation markers (CRP, IL-6), anthropometrics, and functional tests. Models built via weighted scores, regression, and ML ensembles across 13–17 markers. Some include grip strength, gait speed, and other physical performance measures.
Common models: InsideTracker InnerAge 2.0, Thorne Biological Age Health Panel
Transcriptomic
Transcriptomic
Measure which genes are actively turned on or off in your cells. Sensitive to immune and metabolic changes, but also to short-term disruptions.
What it measures
The activity level of your genes — which ones are ramped up and which are dialed down. This activity pattern shifts with age and reflects your immune and metabolic state.
Strengths
- +Sensitive to changes in immune and metabolic health
- +Can detect responses to recent physiological shifts
- +Well-established in research settings
- +Large-scale models exist in published literature
Limitations
- −Very reactive to short-term events (infections, stress) — hard to separate signal from noise
- −Results depend heavily on sample processing
- −Mostly available as part of broader multi-omic panels, not standalone
- −Results from one lab platform may not translate to another
Samples:BloodStoolTissue biopsy
Best for: Research-grade aging studies and monitoring immune or metabolic health as part of a broader panel
Technical details›
Measures gene expression levels via bulk RNA-seq, microarrays, or single-cell approaches. Models use linear regression, elastic net, ML ensembles, and deep learning. Metatranscriptomic stool/blood models report moderate R². Cell-type deconvolution and batch-effect correction are critical preprocessing steps. Cross-platform portability remains limited.
Common models: Research models using blood and stool gene-expression data
No consumer tests indexed yet
Proteomic
Proteomic
Measure hundreds of proteins in your blood to predict disease risk and identify which organs may be aging faster than others.
What it measures
Panels of proteins circulating in your blood. Different organs release different proteins, so these panels can reveal not just overall aging but which parts of your body may be aging faster or slower.
Strengths
- +Strong links to disease risk and mortality — close to what your body is actually doing
- +Large population studies show good prediction of health outcomes
- +Can reveal organ-level aging patterns (brain, heart, liver, etc.)
- +Adds meaningful information beyond just knowing your calendar age
Limitations
- −Results depend on which measurement platform was used
- −Results from one platform may not match another
- −Still expensive for consumer use
- −Mostly available through clinical or research partnerships
Samples:Venous blood
Best for: Understanding disease and mortality risk, learning which organs may be aging faster, clinical research
Technical details›
Measures panels of plasma proteins via aptamer-based (SomaScan, ~7,000 proteins) or antibody-based (Olink, ~3,000 proteins) platforms. Models built with elastic net / penalized regression and ML ensembles on large biobank cohorts. Organ-specific aging signatures derived by training on organ-enriched protein sets. Platform dependence is a key limitation — aptamer and antibody panels have limited overlap.
Common models: Large-cohort proteomic aging models from biobank studies
No consumer tests indexed yet
Microbiome
Microbiome
Profile the bacteria and other microbes in your gut or mouth to estimate biological age. Responds quickly to diet and lifestyle, but results vary a lot between populations.
What it measures
The composition of microbes living in your gut (or mouth). The balance of bacterial species shifts with age, and models use these patterns to estimate how old your microbiome looks.
Strengths
- +Can estimate age from microbial patterns
- +Exploratory connections to diet and disease
- +Responds noticeably to lifestyle changes
- +Non-invasive sample collection (stool or oral swab)
Limitations
- −Heavily influenced by where you live, what you eat, and recent antibiotic use
- −Models trained on one population may not work well for another
- −Many microbiome tests offer age-like insights without rigorous clock validation
- −Day-to-day fluctuations can be large
Samples:StoolOral swab
Best for: Exploratory gut health monitoring, tracking response to dietary changes, research applications
Technical details›
Measures microbial taxonomic and functional profiles via 16S rRNA sequencing, shotgun metagenomics, or metatranscriptomics. Models use ML ensembles and deep learning. Age prediction accuracy is moderate (typical MAE 5–8 years). Portability across geographic populations is a major unsolved challenge due to strong confounding by diet, geography, and antibiotic history.
Common models: ML and deep learning models trained on large gut microbiome datasets
Multi-omic
Composite / Multi-Omic
Combine multiple types of biological data — DNA tags, proteins, metabolites, and blood markers — for a potentially more complete picture of aging.
What it measures
Multiple biological layers at once, aiming to capture different dimensions of aging in a single score. Think of it as running several clock types together and combining their answers.
Strengths
- +Broader coverage — measures aging across multiple biological layers
- +Can combine the strengths of individual clock types
- +May capture the complexity of aging better than any single approach
Limitations
- −Harder to understand what is driving your score
- −Combining different data sources introduces its own errors
- −Many proprietary composites lack independent testing by outside researchers
- −Higher cost because multiple lab tests are needed
Samples:BloodMultiple sample types
Best for: People who want the most comprehensive aging assessment available, and research settings with multiple data streams
Technical details›
Uses stacked models, ensembles, or deep learning to integrate multiple omics layers (epigenomic, proteomic, metabolomic) with clinical markers. Dataset shift risk increases when combining platforms with different calibration schemes. Independent benchmarking of proprietary composites remains limited.
Common models: TruDiagnostic multi-report packages, Viome multi-biosource approach
No consumer tests indexed yet
Immune
Immune / Inflammatory
Measure the composition and activity of your immune system to estimate your immune age or inflammatory age. Directly relevant to how well your body fights disease.
What it measures
The types of immune cells in your blood, the signaling molecules they produce, and overall immune system balance. These patterns predict how well your immune system is holding up with age.
Strengths
- +Predicts immune health, mortality risk, and frailty
- +Immune aging trajectories predict mortality beyond standard risk factors
- +Grounded in established immunology research
- +Directly measures the chronic low-grade inflammation linked to aging
Limitations
- −Mostly available through clinics or partnerships, not direct-to-consumer
- −Pricing is often not publicly listed
- −Fewer consumer-focused validation studies
- −Requires a standard blood draw
Samples:Venous blood
Best for: Monitoring immune health, tracking inflammation, and clinical aging research
Technical details›
Measures immune cell subset composition (flow cytometry), cytokines/chemokines (multiplex assays), or multi-omic immunome profiles. Derived from frameworks like the Stanford 1,000 Immunomes project. IMM-AGE trajectory models use deep learning to predict all-cause mortality beyond standard risk factors. iAge is a composite inflammatory aging score trained on cytokine/chemokine panels.
Common models: Edifice Health iAge (Inflammatory Age)
Telomere
Telomere-Based
Measure the protective caps on the ends of your chromosomes. These caps shorten as cells divide, but as a standalone aging clock, telomere length has significant limitations.
What it measures
The length of telomeres — protective end-caps on your chromosomes that shorten each time a cell divides. Shorter telomeres are associated with aging, but the relationship is noisy and many experts consider this a weak standalone aging indicator.
Strengths
- +Measures a well-understood aspect of cellular aging
- +One of the longest-studied aging biomarkers
- +Can complement other clock types as supporting data
Limitations
- −Measurements vary a lot depending on the method used
- −Only reflects aging in the specific cells tested, not the whole body
- −Most experts consider it a weak standalone aging clock
- −Less accurate than multi-biomarker approaches for general aging prediction
Samples:Blood
Best for: People interested in telomere biology specifically, or as supplementary data alongside other clocks
Technical details›
Measures leukocyte telomere length via qPCR (most common, ratio-based), flow-FISH (single-cell resolution), or Terminal Restriction Fragment (TRF, gold standard but labor-intensive). Results typically reported as T/S ratio or absolute length in kilobases. High inter-assay variability, especially with qPCR. Some epigenetic vendors include telomere length estimates as auxiliary reports derived from methylation data rather than direct measurement.
Common models: Life Length diagnostics, telomere add-ons from epigenetic testing providers
No consumer tests indexed yet
Imaging
Imaging-Based
Use AI analysis of medical images — facial photos, chest X-rays, or brain scans — to estimate biological age. Promising in clinical research, but not yet a consumer product.
What it measures
Visual patterns in medical images that change with age. AI models learn to spot aging-related features in photos and scans that human eyes might miss, and estimate how old the body (or a specific organ) appears.
Strengths
- +Shown to predict health outcomes in specific medical contexts
- +Some models validated in large clinical studies
- +Non-invasive, especially facial photo approaches
Limitations
- −Privacy and ethics concerns, especially for facial analysis
- −Requires careful testing for demographic bias
- −Not available as standalone consumer tests
- −Measures aging in one organ or area, not whole-body aging
Samples:Facial photoChest X-rayBrain MRI
Best for: Clinical research and organ-specific prognostic assessment — not yet consumer-ready
Technical details›
Uses convolutional neural networks (CNNs) and deep learning applied to medical images. FaceAge extracts aging features from facial photos and has shown prognostic value in oncology cohorts. CXR-Age predicts mortality from chest radiographs. Brain MRI models estimate brain age gap (predicted minus chronological age). All require careful bias auditing across demographic groups.
Common models: FaceAge (used in cancer research), CXR-Age (chest X-ray mortality prediction)
No consumer tests indexed yet