14 evidence-first reviews in this topic, built on 100 peer-reviewed citations. Null and negative trials are reported alongside the positive ones — that is the point of the format.
Epigenetic clocks predict mortality consistently across populations and unreliably within one person: technical noise alone produced replicate deviations of up to 9 years in six standard…
Where the evidence lands: Epigenetic age acceleration predicts all-cause mortality at population scale: across 13 cohorts and 13,089 people every measure was predictive independently of chronological age (P<=8.2e-9), and a review of 23 studies and 41,607 participants estimated an 8-15% higher mortality risk per 5-year increase, while flagging publication bias and calling the disease and longevity evidence inconclusive. Generation matters. In 490 TILDA participants, Horvath and Hannum acceleration predicted nothing; GrimAge acceleration predicted 8 of 9 outcomes before adjustment and 4 of 9 after. In ESTHER, per-SD mortality hazard ratios were 1.32 for PhenoAge and 1.47 for GrimAge, but adding both clocks to a model with age, sex, leukocyte composition, an 8-CpG score and a frailty index moved Harrell's C only from 0.759 to 0.766. Reliability is the weak point: technical noise produced up to 9 years of deviation between replicates in six clocks, narrowed to within 1.5 years for most replicates. Intervention evidence is thin: 2 years of 25% caloric restriction (CALERIE, n=220) slowed DunedinPACE but not PhenoAge or GrimAge, with small effect sizes; in MACRO (n=144), changes in epigenetic aging did not track cardiometabolic change. Telomere tests reverse direction by outcome: the longest decile carried 26% higher cancer incidence in 26,540 Chinese adults.
The Omega-3 Index promises to turn fish-oil status into a single percentage you can track. The biomarker is scientifically serious — and the leap from a higher index to a longer life…
Where the evidence lands: A higher Omega-3 Index is consistently associated with lower all-cause and cardiovascular mortality across dozens of cohorts (roughly 30% lower fatal CHD risk near 8% versus 4%), but that is association — large supplementation trials were neutral, and one 2024 analysis flagged higher atrial fibrillation and stroke risk with fish-oil use.
Because every atherogenic particle carries exactly one apoB molecule, an apoB test counts plaque-forming particles directly — and in a 233,455-person pooling it beat LDL-cholesterol by…
Where the evidence lands: ApoB refines cardiovascular risk especially in diabetes, obesity, or high triglycerides, where LDL-C misleads; it is a risk marker rather than a treatment, and not yet recommended as routine universal screening.
Cortisol patterns track mortality and cardiovascular risk across large cohorts, and late-night salivary cortisol screens for Cushing syndrome at roughly 92% sensitivity and 96% specificity…
Where the evidence lands: Cortisol is a genuine research biomarker with a narrow validated clinical use; as an individual 'stress damage' diagnostic, a consumer saliva test is weak.
Serum ferritin is the standard window into the body's iron stores — and iron status bends both ways: deficiency drags down energy and resilience, while overload tracks with diabetes and…
Where the evidence lands: Ferritin's clearest value is diagnosing iron deficiency and iron overload; within the normal range it was not associated with mortality in a large national cohort, and because inflammation itself raises ferritin, every reading needs context.
HOMA-IR turns one fasting blood draw into an index of insulin resistance — validated against the clamp, predictive of diabetes and cardiovascular disease in large cohorts, and yet…
Where the evidence lands: Fasting insulin and HOMA-IR are useful, validated surrogate markers of insulin resistance; they are not a diagnosis, they have no universal threshold, and some important associations weaken or vanish with age.
Lp(a) is 80–90% genetically determined, elevated in roughly one in five people, and causally linked to heart attack, stroke, and aortic valve disease — yet it is absent from standard…
Where the evidence lands: Testing Lp(a) once is well supported for risk stratification; because no Lp(a)-lowering drug is yet approved, a high result currently guides aggressive management of everything else rather than a specific treatment.
Largely independent of muscle mass, diet, age, and sex, cystatin C-based eGFR flags more chronic kidney disease than creatinine alone (13.7% vs 9.7% below 60) and predicts death and…
Where the evidence lands: Cystatin C is the confirmatory kidney measure current guidelines recommend combining with creatinine; it sharpens risk prediction but is not flawless — and it is a biomarker, not a product endorsement.
GGT is a cheap, old liver enzyme test that keeps showing up in mortality studies — even within the normal range. The associations are real and consistent; whether the number causes…
Where the evidence lands: Higher serum GGT, even inside the reference range, is consistently associated with higher all-cause and cardiovascular mortality and several chronic diseases — but the evidence is observational, a Mendelian-randomization study found no causal effect on diabetes, and the test is too nonspecific to interpret alone.
High-sensitivity CRP measures the faint background inflammation that tracks with heart attack, stroke, frailty, and dementia — and two landmark trials proved the risk it flags is…
Where the evidence lands: hs-CRP is a validated marker of residual inflammatory cardiovascular risk — useful, non-specific, and best repeated when you are well — but Mendelian randomization indicates CRP itself is probably not the culprit, and for most people it adds only modest predictive value beyond traditional risk factors.
Across cohorts totaling millions, weaker grip strength tracks higher all-cause and cardiovascular mortality as well as or better than systolic blood pressure — a powerful marker of…
Where the evidence lands: Grip strength is a cheap, well-normed vital sign of aging; know your number against age- and sex-specific ranges, but treat it as a prompt for overall strength and activity, not a target in itself.
CGMs reveal biological variation, but a visible glucose spike is not automatically harmful and consumer feedback has not been shown to improve long-term outcomes in healthy people.
Where the evidence lands: A CGM trace is not a direct health score. Healthy people show variable post-meal responses, sensor and plasma values can disagree, and outcome benefits from consumer CGM use remain unproven.
The headline is based on a higher-risk subgroup, not the trial population as a whole. Hearing care still delivered a large and durable communication benefit.
Where the evidence lands: ACHIEVE found no cognitive benefit in its full 977-person trial population. The widely quoted 48% reduction came from a prespecified higher-risk subgroup; hearing-related communication improved substantially and durably across the trial.
The ApoB Home Test Kit measures apolipoprotein B, which may be a more accurate indicator of cardiovascular risk than traditional cholesterol tests.
Where the evidence lands: The ApoB Home Test Kit, by measuring apolipoprotein B particle count, offers a potentially more accurate way to assess cardiovascular risk compared to traditional cholesterol tests, but it should be used as part of a broader health management strategy and in consultation with a healthcare professional.
97 distinct peer-reviewed papers, published 1985–2026, across 71 journals. Each links to its PubMed record; where Magellan has published a full study write-up, that is linked too.
Measurement tool
Large Mendelian randomization studies, prospective cohorts and meta-analyses — in one pooling of…
Prefer the interactive version? Open the article feed in the Magellan app →