PhenoAge Scientific Foundation

PhenoAge is designed to calculate an individual’s biological age based on various biometric inputs. It provides insights into a person’s health status by comparing their biological age to their chronological age. The algorithm is based on peer-reviewed scientific research by Dr. Morgan Levine and colleagues. It has been validated across diverse populations and is widely used in clinical research, population studies, and preventive medicine.
For practical use cases, read more about our PhenoAge Calculator. For API documentation and technical details, refer to the PhenoAge Calculator API. Or read about our entire Biometric Calculator suite.
Key Publications
Levine, M. E. et al. (Apr 2018). “An epigenetic biomarker of aging for lifespan and healthspan.” Aging, 10(4), 573-591. https://doi.org/10.18632/aging.101414
PhenoAge was used as the target for training a DNA methylation-based biomarker. The authors first developed PhenoAge using NHANES III data and a weighted combination of nine blood biomarkers and chronological age, selected for their strong association with mortality risk. This phenotypic measure was then regressed on DNA methylation data from the InCHIANTI and WHI cohorts to identify CpG sites whose methylation patterns approximate biological aging across tissues. A supplement provides the precise formula, biomarker list, and model coefficients used to calculate PhenoAge.
Levine, M. E. et al. (Dec 2018). “A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: A cohort study.” PLOS Medicine, 15(12), e1002718. https://doi.org/10.1371/journal.pmed.1002718
PhenoAge is a measure based on clinical biomarkers that estimates a person’s biological age in relation to their mortality risk, providing insight into their rate of aging. In this study, researchers assessed its usefulness across diverse populations using data from over 11,000 adults in the NHANES IV survey. PhenoAge was strongly associated with all-cause and cause-specific mortality, even after adjusting for chronological age, and was predictive across subgroups, including healthy individuals and the oldest-old. The findings support its potential as a clinical and research tool for identifying at-risk individuals and evaluating aging-related interventions, though more research in other cohorts is needed.
Levine, M. E. et al. (Feb 2019). “Correction: A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: A cohort study.” PLOS Medicine, 16(2), e1002728. https://doi.org/10.1371/journal.pmed.1002760
A correction was issued for a study on the PhenoAge measure, which estimates biological aging and predicts health risks across diverse U.S. populations. The original article omitted a crucial step in the equation used to calculate PhenoAge, making it unsolvable as published. The corrected formula now includes the proper transformation of clinical variables, such as albumin, glucose, and white blood cell count, alongside chronological age. This clarification ensures accurate computation of PhenoAge for research and clinical applications in aging and mortality risk assessment.
The PhenoAge formula was first introduced in April 2018, using NHANES III data to create a formula (PhenoAge1) that combined nine blood biomarkers and chronological age to estimate mortality risk, which was then converted with specific constants into a biological age estimate. In December 2018, a second paper presented a revised formula (PhenoAge2) based on NHANES IV data, using slightly different constants but inadvertently omitting the intermediate step that converts mortality risk to biological age, rendering the formula incorrect as published. The error was corrected in February 2019, when the missing step was restored and the final formula (corrected PhenoAge2) was published. These updates mark the progression from the original PhenoAge formula to a corrected and more broadly validated version.
Popular Adjustment
Cramer, J. G. (2018). “New blood tests can reveal your life expectancy.” Age Reversal Forum. View Discussion
After the PhenoAge formula was published, researchers noticed that it tended to overestimate biological age for very old individuals, likely due to limited data in that age range. To address this, an informal adjustment (adjusted PhenoAge) was proposed by fitting a curve to Levine’s original graph, helping to bring estimates more in line with DNA methylation-based aging patterns. Although not part of the official publications, this adjustment became popular for improving accuracy, especially in older adults. It is often used to refine clinical PhenoAge scores into more realistic biological age estimates.
Required Biomarkers
PhenoAge requires the following clinical biomarkers for calculation:
- Age (chronological age in years)
- Albumin (g/dL)
- Alkaline Phosphatase (ALP) (U/L)
- Creatinine (mg/dL)
- C-Reactive Protein (CRP) (mg/L)
- Glucose (mg/dL)
- Lymphocyte Percentage (%)
- Mean Corpuscular Volume (MCV) (fL)
- Red Blood Cell Distribution Width (RDW) (%)
- White Blood Cell Count (WBC) (K/μL)
All biomarkers are required for accurate PhenoAge calculation. CRP values are log-transformed during the calculation process.
Algorithm Overview
PhenoAge uses a Cox proportional hazards regression model to predict mortality risk based on clinical biomarkers. The algorithm:
- Converts biomarkers to standardized units
- Applies weighted coefficients to each biomarker
- Calculates a mortality score using the weighted sum
- Converts the mortality score to biological age using a mathematical transformation
- Applies an adjustment formula to correct for age-related bias
The final result provides both the raw PhenoAge and an adjusted version that accounts for age-related systematic errors.
Validation and Limitations
PhenoAge has been validated in multiple studies:
- NHANES IV cohort: 11,000+ adults with mortality follow-up
- Diverse populations: Validated across different ethnicities and age groups
- Clinical outcomes: Predicts all-cause and cause-specific mortality
Current Limitations:
- Limited data for very old individuals (>85 years)
- May not capture all aspects of biological aging
- Requires standardized laboratory measurements
- Results should be interpreted in clinical context
The algorithm performs best in middle-aged to older adults and should be used alongside other clinical assessments.
Technical Implementation
Our PhenoAge implementation includes:
- Two-step calculation: Raw PhenoAge and adjusted PhenoAge
- Log transformation: CRP values are log-transformed before weighting
- Age adjustment: Correction formula reduces age-related bias
- Validation checks: Negative biological ages are rejected
- Legacy support: Accepts both standard and legacy parameter names
The algorithm returns both PhenoAge1 and PhenoAge2 variants, with the adjusted PhenoAge2 being the recommended final result.
Usage Guidelines
When to use PhenoAge:
- Adults aged 20-85 years
- Standard clinical laboratory results available
- Assessment of biological aging and mortality risk
- Research studies of aging interventions
Interpretation:
- PhenoAge < Chronological Age: Younger biological age
- PhenoAge > Chronological Age: Older biological age
- Difference: Years of accelerated or decelerated aging
Clinical Considerations:
- Results should be interpreted by healthcare professionals
- Consider alongside other clinical assessments
- Not a diagnostic tool for specific diseases
- May vary with acute illness or inflammation