BioMIR® is a privacy-first, on-device BioClock that converts Apple Health-authorized data into an interpretable biological-age estimate expressed in Δ-years.
Instead of asking users to manually enter biomarkers, BioMIR analyzes signals they already generate through Apple Health-connected wearables, nutrition apps, recovery tools, blood-pressure monitors, glucose meters, smart scales, and clinical records. The goal is longitudinal visibility: a clearer view of whether measurable physiologic, behavioral, and cardiometabolic patterns are moving toward aging acceleration, stabilization, or deceleration.
BioMIR does not replace clinical care. It provides educational biological-aging context between clinic visits using deterministic, interpretable modeling.
THREE MODELING DOMAINS
Allostatic Load
HRV, resting heart rate, VO₂ max, total sleep, and deep sleep provide context for autonomic tone, recovery, physiologic load, and reserve capacity.
Behavioral Adaptation
Activity, meditation, daylight exposure, alcohol, sodium, energy intake, and total carbohydrate intake provide behavioral and environmental context. Total carbohydrate is treated as a dietary-load proxy, not a nutrition score.
Cardiometabolic Anchors
Systolic blood pressure, fasting glucose, body weight, and derived BMI provide vascular, glycemic, and body-composition context.
BioMIR models these domains separately because adaptive behavior, physiologic recovery, and cardiometabolic status are related but not interchangeable.
BIOMARKER FEEDBACK
BioMIR ranks the inputs contributing most to the selected Δ-years signal and explains why they matter. Domain and biomarker cards may include interpretation boundaries, evidence context, contributor ranking, and a Top Action: an educational prompt linking the selected input to a relevant behavior or measurement pattern.
Where available, biomarker cards include published literature and meta-analytic sources used for model design, biological rationale, hazard-gradient weighting, and longevity-context interpretation.
WHAT YOU CAN EXPLORE
Review biological-aging patterns across wearable, behavioral, cardiometabolic, and compatible clinical inputs.
Explore daily, weekly, monthly, quarterly, and yearly windows.
Compare mean, median, sample size, confidence-interval bands, and rate-of-change summaries.
Use biomarker cards to understand each input’s biological relevance and associated Top Action.
Add KDM and Levine PhenoAge context when compatible clinical biomarkers are available.
Explore Demo Data and simulations without changing live data.
MODELING APPROACH
BioMIR is deterministic, interpretable, and mechanistically structured. It uses population-scale reference data, age- and sex-aware normalization, asymmetric scaling, hazard-gradient weighting, age regression, smoothing, and contributor ranking.
Hazard-gradient weighting gives greater influence to biomarkers with stronger published associations across mortality, morbidity, physiologic burden, and longevity literature. Age regression maps deviations from reference expectations onto a chronological-age scale to estimate Δ-years.
PRIVACY-FIRST ARCHITECTURE
BioMIR runs on device. Authorized health and clinical data are processed locally within the app sandbox and are not uploaded to external servers.
BioMIR accesses health and clinical data only with explicit authorization. It does not use health or clinical data for ads, tracking, profiling, sale, machine-learning training, or research datasets.
IMPORTANT LIMITATIONS
BioMIR is informational and educational only. It does not diagnose, treat, predict, manage, cure, mitigate, or prevent disease. It does not provide medical advice, triage, prognosis, disease monitoring, clinical decision support, or replacement for medical care. Consult a qualified clinician before making medical decisions.
Privacy: https://biomir.github.io/biomir-privacy/
EULA: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/
Show more
Show less
