BioMIR

BioMIR® offers scientific visibility into biological-aging dynamics without reducing health data to a wellness score. BioMIR converts wearable data, behavior logs, biosensor inputs, cardiometabolic measures, and clinical bloodwork into an on-device biological-age signal. The output is Δ-years: biomarker deviation expressed on a chronological-age scale using age- and sex-aware reference context. Routine labs are episodic, while wearables and home monitors collect frequent signals that most platforms compress into simplified scores. BioMIR bridges them: self-tracking contextualized with clinical biological-age reference models when labs are available. THREE MODELING DOMAINS Allostatic Load HRV, resting heart rate, VO₂ max, total sleep, and deep sleep provide physiologic context for autonomic tone, recovery, burden, and reserve. Behavioral Adaptation Activity, meditation, light 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. These domains are modeled separately because recovery, behavior, and cardiometabolic status are related but not interchangeable. BIOMARKER FEEDBACK BioMIR identifies contributors most responsible for the selected Δ-years signal. Each domain and biomarker card may provide ranking, interpretation boundaries, evidence context, and Top Action: an education-focused prompt linking the selected biomarker to a high-priority behavior or measurement context. Where available, cards include published research and meta-analytic sources used for model design, biological rationale, or hazard-weight derivation. WHAT YOU CAN DO See what is driving your biological-aging pattern across wearable, behavioral, cardiometabolic, and clinical inputs. Review daily, weekly, monthly, quarterly, and yearly windows. Use biomarker cards to see why an input matters and which Top Action is associated with it. Compare mean, median, raw data, sample size, confidence-interval bands, and rate-of-change summaries. Add KDM and Levine PhenoAge reference models when clinical biomarker data are available. Explore Demo Data and simulations without changing live data. WHO BioMIR IS FOR BioMIR is for users who want scientifically grounded biological-age modeling without black-box wellness scores, including people tracking healthspan, recovery, behavior change, cardiometabolic patterns, or longitudinal biomarkers. MODELING APPROACH BioMIR computation is deterministic, interpretable, and mechanistically structured. It uses NHANES and other population-scale reference data, age- and sex-aware context, asymmetric scaling, hazard weighting, age regression, smoothing, and contributor ranking. Hazard weighting gives greater influence to signals with stronger published associations to physiological burden or long-term biological risk context. Age regression maps biomarker deviation from reference expectations onto a chronological-age scale, producing the Δ-years estimate. PRIVACY-FIRST ARCHITECTURE BioMIR operates on device. Authorized health and clinical data are processed locally within the app sandbox, not uploaded or backed up to external servers. BioMIR accesses health and clinical data only with explicit authorization, and does not use health or clinical data for ads, tracking, profiling, monetization, sale, brokerage, ML training, or research datasets. IMPORTANT LIMITATIONS BioMIR is informational only. It does not diagnose, treat, predict, monitor, manage, cure, mitigate, or prevent disease; provide medical advice, triage, prognosis, disease monitoring, or clinical decision support; or replace medical care. Privacy: https://biomir.github.io/biomir-privacy/ EULA: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/