Healthy aging
Biomarkers, surrogate endpoints, and actual health outcomes
On this page
Clinical outcomes directly measure how people feel, function, or survive. A surrogate endpoint substitutes for an outcome; its usefulness depends on evidence in a defined context.[1]
What validation changes
FDA describes extensive supporting evidence for validated surrogates. A surrogate considered reasonably likely to predict benefit has less support and may require later confirmation of clinical benefit.[1]
Biomarkers have different jobs
A biomarker is a measured characteristic indicating a biological process, disease process or response. FDA categories include susceptibility or risk, diagnosis, monitoring, prognosis, prediction of treatment response, response to treatment and safety. One measurement may have more than one role, but qualification applies to a defined context of use. Recognizing a marker’s name is not enough: the biological characteristic, specimen and measurement method also need definition.[2]
Interpreting a laboratory number
Laboratory methods, units, and ranges can differ. Medicines, food, and exercise affect some results. Compare measurements in clinical context; a small change alone does not demonstrate treatment benefit.[3]
An approval can still require clinical-outcome confirmation
FDA’s semaglutide MASH approval used improvement in liver biopsy findings under the accelerated pathway. FDA describes an ongoing 240-week study to confirm clinical benefit, including outcomes related to liver disease. Histological improvement and confirmed reduction in death, transplantation or other clinical complications are distinct evidence questions. This uncertainty differs from the direct cardiovascular-event outcome tested in SELECT.[4]
Measurement validation and biomarker purpose answer different questions
| Term | Question it addresses | Important distinction |
|---|---|---|
| Analytical validation | Does the test measure its intended target adequately? | Does not itself establish the clinical meaning.[6] |
| Clinical validation | Does the marker measure or predict the relevant clinical concept? | The concept and intended use must be specified.[6] |
| Prognostic biomarker | What is the likelihood of an event or progression in people with the condition? | Not automatically evidence of a different treatment effect.[7] |
| Predictive biomarker | Does exposure or treatment affect marker-defined people differently? | Different from prognosis alone.[7] |
An accurate measurement and a useful clinical interpretation are separate steps
FDA–NIH BEST separates analytical validation from clinical validation. A reproducible measurement can still need evidence connecting it with the claimed clinical concept. Conversely, interpreting a marker clinically requires a test that adequately measures it. The two questions cannot be collapsed into “validated” without saying what was validated.[6]
Predictive has a specific meaning in this framework
In BEST terminology, prognostic concerns the likelihood of an event in someone with a condition. Predictive concerns a different favorable or unfavorable response to an exposure. A marker associated with poor outcomes is not automatically proof that a particular treatment works better in that marker-defined group.[7]
FDA’s context-of-use description combines the biomarker category with its intended job in drug development. Population, disease stage and development setting may be part of that description. A claim should therefore name the use rather than treating the marker as a universally qualified health score.[8]
Historical examples test a practical assumption
A treatment can change a measured signal in the intended direction while having other effects that alter the final health outcome. Two historical trials illustrate why an intervention's clinical outcomes matter alongside its biomarker results.[9][10]
Evidence & uncertainty
A changed biomarker does not automatically establish improved health or longer life. Some surrogate endpoints are well validated; validation cannot simply be assumed for a new use.[1]
Selected-source overview · Independent review pendingPublished questions
Does a better biomarker always mean better health?
Selected-source overview · Independent review pending. Read the evidence context.
No. Its relation to clinical benefit needs evidence in that context.[1]
Evidence behind this answer
- Surrogate Endpoint Resources
agency explanation · Date not stated
Selected page sections · Clinical endpoints; validation; accelerated approval
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Are all surrogate endpoints useless?
Selected-source overview · Independent review pending. Read the evidence context.
No. Some have extensive evidence supporting their use.[1]
Evidence behind this answer
- Surrogate Endpoint Resources
agency explanation · Date not stated
Selected page sections · Clinical endpoints; validation; accelerated approval
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Does a biological-age score establish longer lifespan?
Selected-source overview · Independent review pending. Read the evidence context.
A score change alone does not demonstrate a survival benefit.[1]
Evidence behind this answer
- Surrogate Endpoint Resources
agency explanation · Date not stated
Selected page sections · Clinical endpoints; validation; accelerated approval
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Is every biomarker a surrogate endpoint?
Selected-source overview · Independent review pending. Read the evidence context.
No. Biomarkers have several roles; substituting for a clinical outcome needs evidence supporting that particular use.[2]
Evidence behind this answer
- About Biomarkers and Qualification
agency explanation · Date not stated
Selected page text; cited sections · Biomarker categories and context of use
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Does FDA qualification make a biomarker valid for every purpose?
Selected-source overview · Independent review pending. Read the evidence context.
No. Qualification concerns a stated context of use, including the defined biological characteristic and measurement context.[2]
Evidence behind this answer
- About Biomarkers and Qualification
agency explanation · Date not stated
Selected page text; cited sections · Biomarker categories and context of use
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Are validated and reasonably likely surrogates identical evidence categories?
Selected-source overview · Independent review pending. Read the evidence context.
No. FDA describes less supporting evidence for reasonably likely surrogates, with clinical confirmation potentially needed.[1]
Evidence behind this answer
- Surrogate Endpoint Resources
agency explanation · Date not stated
Selected page sections · Clinical endpoints; validation; accelerated approval
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Can I directly compare small changes from different laboratories?
Selected-source overview · Independent review pending. Read the evidence context.
Methods, units, ranges, and clinical context can complicate comparisons.[3]
Evidence behind this answer
- How to Understand Your Lab Results
agency explanation · Date not stated
Selected page text; cited sections · Reference ranges, units, methods and interpretation
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Can a medicine be approved while an important outcome question remains?
Selected-source overview · Independent review pending. Read the evidence context.
Yes. Accelerated approval can use a surrogate considered reasonably likely to predict benefit, with subsequent clinical confirmation required.[4]
Evidence behind this answer
- FDA Approves Treatment for Serious Liver Disease Known as MASH
agency explanation · 2025
Selected page text; cited sections · Accelerated approval, biopsy findings and confirmatory follow-up
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
If a wellness product improves a blood marker, does that mean it improves health?
Selected-source overview · Independent review pending. Read the evidence context.
Not automatically. A biomarker measures biology; patient-important outcomes include symptoms, functioning and survival. A validated surrogate needs evidence that changing it reliably predicts a specific clinical benefit in its defined context. A correlation, plausible mechanism or marker improvement alone cannot establish that a wellness product causes better health, and harms still matter.[2][5]
Evidence behind this answer
- About Biomarkers and Qualification
agency explanation · Date not stated
Selected page text; cited sections · Biomarker categories and context of use - Validated Surrogate Endpoint — BEST Resource
agency methodology · Date not stated
Indexed original-resource passages inspected; direct retrieval returned a verification page · Definition; clinical-context requirement; correlation versus trial-level validation
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Does analytical validation automatically establish clinical validation?
Selected-source overview · Independent review pending. Read the evidence context.
No. Measuring the intended target and connecting it to a relevant clinical concept are separate validation questions in the BEST framework.[6]
Evidence behind this answer
- Validation — BEST Resource
agency methodological definitions · Date not stated
Indexed original definition passages inspected; direct page returned an access check · Indexed original definitions of analytical and clinical validation
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Does a poor-prognosis biomarker prove a treatment works better in that group?
Selected-source overview · Independent review pending. Read the evidence context.
No. Prognosis and differential treatment response are different concepts. A predictive claim needs evidence for the relevant exposure comparison.[7]
Evidence behind this answer
- Understanding Prognostic versus Predictive Biomarkers — BEST Resource
agency methodological definitions · 2016
Indexed original introduction inspected; direct page returned an access check · Original introduction: prognosis versus differential exposure response
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Is saying validated biomarker enough without naming its intended use?
Selected-source overview · Independent review pending. Read the evidence context.
No. FDA describes context of use through the biomarker category and its intended job, with relevant population and setting.[8]
Evidence behind this answer
- Context of Use
agency qualification framework · Date not stated
Selected relevant official page passages inspected · Category plus intended drug-development use and population/setting description
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
Are predictive and prognostic biomarkers interchangeable terms?
Selected-source overview · Independent review pending. Read the evidence context.
No. Prognostic concerns event likelihood in a condition; predictive concerns differences in the response to a treatment or other exposure.[7]
Evidence behind this answer
- Understanding Prognostic versus Predictive Biomarkers — BEST Resource
agency methodological definitions · 2016
Indexed original introduction inspected; direct page returned an access check · Original introduction: prognosis versus differential exposure response
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
If a drug improves a marker but harms people, did the marker cause the harm?
Selected-source overview · Independent review pending. Read the evidence context.
That does not follow. A drug can affect several pathways. ILLUMINATE showed that the overall intervention worsened outcomes, but it did not establish that raising HDL itself caused the excess harm.[10]
Evidence behind this answer
- Effects of Torcetrapib in Patients at High Risk for Coronary Events
Randomized trial · 2007
Original full text read through Firecrawl · Original abstract; outcomes; discussion of uncertain mechanisms and off-target effects
This answer summarizes selected sources. An independent appraisal of the complete evidence has not been completed.
A useful marker is not automatically a validated surrogate
A marker can correlate with disease without showing that changing it improves how people feel, function or survive. Validation needs supporting evidence for its intended treatment and outcome context. FDA distinguishes well-supported surrogates from those reasonably likely to predict benefit, used with greater uncertainty and possible requirements for confirmation. The same marker can be informative in one setting and insufficient in another.[1]
Suppressing an abnormal rhythm signal did not ensure survival
CAST enrolled selected people after a heart attack whose extra beats from the heart's lower chambers could be suppressed by encainide or flecainide. In the randomized comparison, treatment increased deaths despite suppressing that rhythm signal. The report did not establish the mechanism of excess mortality. This finding concerned particular drugs and a post-heart-attack population; it is not a claim about every arrhythmia or antiarrhythmic medicine.[9]
Improved lipid markers accompanied worse outcomes
ILLUMINATE compared torcetrapib plus atorvastatin with atorvastatin alone in high-risk patients. HDL cholesterol rose and LDL cholesterol fell, yet cardiovascular events and deaths increased, and the trial stopped early. Blood pressure and other physiological changes also occurred. The investigators considered off-target effects possible, while the mechanism remained uncertain. The trial did not isolate the HDL change as the cause of harm.[10]
Questions still to cover
Further work: individual aging-test validation and reproducibility; full drug-specific surrogate validation and confirmatory outcomes; measurement error and consequences of repeated low-risk testing.
References
- Surrogate Endpoint Resources
Agency explanation
FDA · Content/publication date: not stated
Accessed: 2026-10-06 UTC · Selected page sections
Cited sections: Clinical endpoints; validation; accelerated approval - About Biomarkers and Qualification
Agency explanation
FDA · Content/publication date: not stated
Accessed: 2026-10-06 UTC · Selected page text; cited sections
Cited sections: Biomarker categories and context of use - How to Understand Your Lab Results
Agency explanation
MedlinePlus · Content/publication date: not stated
Accessed: 2026-10-06 UTC · Selected page text; cited sections
Cited sections: Reference ranges, units, methods and interpretation - FDA Approves Treatment for Serious Liver Disease Known as MASH
Agency explanation
FDA · Content/publication date: 2025
Accessed: 2026-10-06 UTC · Selected page text; cited sections
Cited sections: Accelerated approval, biopsy findings and confirmatory follow-upSelected approval explanation; ongoing clinical-outcome confirmation described by FDA. Complete prescribing information not appraised.
- Validated Surrogate Endpoint — BEST Resource
Agency methodology
FDA–NIH Biomarker Working Group · Content/publication date: not stated
Accessed: 2026-10-06 UTC · Indexed original-resource passages inspected; direct retrieval returned a verification page
Cited sections: Definition; clinical-context requirement; correlation versus trial-level validationSelected original-resource excerpts only; no full chapter appraisal. No specific wellness-product validation is claimed.
- Validation — BEST Resource
Agency methodological definitions
FDA–NIH Biomarker Working Group · Content/publication date: not stated
Accessed: 2026-10-07 UTC · Indexed original definition passages inspected; direct page returned an access check
Cited sections: Indexed original definitions of analytical and clinical validationSelected definitions only; no full independent biomarker-validation appraisal.
- Understanding Prognostic versus Predictive Biomarkers — BEST Resource
Agency methodological definitions
FDA–NIH Biomarker Working Group · Content/publication date: 2016
Accessed: 2026-10-07 UTC · Indexed original introduction inspected; direct page returned an access check
Cited sections: Original introduction: prognosis versus differential exposure response - Context of Use
Agency qualification framework
US FDA · Content/publication date: not stated
Accessed: 2026-10-07 UTC · Selected relevant official page passages inspected
Cited sections: Category plus intended drug-development use and population/setting descriptionDrug-development qualification context, not validation of every consumer health score.
- Mortality and Morbidity in Patients Receiving Encainide, Flecainide, or Placebo — The Cardiac Arrhythmia Suppression Trial
Randomized trial
Echt and colleagues; CAST investigators · Content/publication date: 1991
Accessed: 2026-10-20 UTC · Original full text read through Firecrawl after native block
Cited sections: Original abstract; randomized population and outcomes; mechanism uncertaintyHistorical selected post-infarction population; harmful intervention does not identify the causal mediator.
- Effects of Torcetrapib in Patients at High Risk for Coronary Events
Randomized trial
Barter and colleagues; ILLUMINATE investigators · Content/publication date: 2007
Accessed: 2026-10-20 UTC · Original full text read through Firecrawl
Cited sections: Original abstract; outcomes; discussion of uncertain mechanisms and off-target effectsClinical harm despite favorable lipid changes; HDL change was not isolated as a causal mediator.
Scope & review
- Population
- People interpreting research or tracking
- Exposure or intervention
- Endpoint selection
- Comparison
- Biomarkers and clinical outcomes
- Outcomes
- Symptoms, function, survival
- Timeframe
- Context-specific evidence
- Jurisdiction
- US FDA framework
Selected-source educational explainer. Cited page sections or indexed abstracts were inspected as specified for each source. This is not a systematic search, a formal evidence grade, or an independent clinical expert review; topic and applicability judgments are editorial. The edition 3.8 check applies to selected additions and named reused passages; older references retain their individual access dates.
Sources checked on their individual access dates. Selected sources were checked; ongoing surveillance and qualified review have separate statuses on the coverage page.
Revision history
- 2026-10-06 UTC: Bounded foundation added with source locators, scope, and unresolved expansion areas.
- 2026-10-06 UTC: Expanded definitions, categories, evidence boundaries and locally sourced questions; selected sources checked with access limits recorded. Independent expert review remains incomplete.
- 2026-10-06 UTC: Added related reading and made remaining editorial scope explicit in edition 3.1; source-access dates retained.
- 2026-10-07 UTC: Edition 3.8: added definitions, worked distinctions and sourced questions about evidence, testing, medicines and exposure. Qualified review remains pending.
- 2026-10-20 UTC: Added selected-source foundational, practical and evidence explanations after reading full target and neighboring guide bodies. Existing body, question and source prefixes preserved; qualified review pending.
What this article still leaves open
- individual aging-test validation and reproducibility
- full drug-specific surrogate validation and confirmatory outcomes
- measurement error and consequences of repeated low-risk testing
- Full source-to-claim appraisal and purpose-specific performance, decision benefit and patient outcomes.
Suggest a correction
Submit a content correction to the editorial queue, or download a local note. Leave out personal medical details. Submitted text is stored privately with a browser receipt session and is visible to this browser and the editor. Clearing cookies removes access to your submission history. This is not an account or a medical service.
Your submitted corrections for this article
Explore the topic map
Browse the health domains connected to this article. The map also identifies topics whose articles remain unwritten.