Refeed Protocols: What the Research Supports (and What It Doesn't)
Carbohydrate loading for performance has strong evidence. Refeeds as a treatment for RED-S do not. Here's what the research actually supports, and what remains experimental.
Engineering Blog
How we build Omnio — health data infrastructure, MCP servers, time-series analytics, and the AI layer that ties it together.
Confounders That Look Like RED-S: Cycle Phase, Illness, Alcohol, Training Spikes
Luteal phase, viral illness, alcohol, heavy training blocks, travel, and sleep debt produce biomarker signatures almost identical to underfueling. Here's how to tell them apart.
Biomarker Signatures of Underfueling: Sleep, HRV, RHR, and Their Limits
The physiological signatures of low energy availability show up in HRV, resting heart rate, sleep, body mass, and menstrual function. Each signal is real and each has significant limits.
energy-availability red-s calculation measurement-error fat-free-mass exercise-energy-expenditure July 30, 2026
How Energy Availability Is Calculated (and Why It's Harder Than It Looks)
Energy availability looks like a simple subtraction, but each input (intake, exercise expenditure, fat-free mass) has 10–30 percent measurement error. Here's what that means.
How AI Reads a Meal Photo: Dual-Model Vision, Perceptual Hashing, and Confidence
A modern meal-photo pipeline has four stages — quality, duplicate detection, LLM vision, and confidence-gated review — plus an enrichment layer. Here's how each step keeps the output honest.
Tracking 35 Micronutrients: Catching Deficiencies Before They Become Symptoms
Macros cover four numbers. The other 30+ micronutrients move independently and most people are short on several. Here's how 35-nutrient tracking surfaces gaps before blood work does.
Your Training Split vs Your Actual Training Split: Schedule Detection from Behavior
The program says upper/lower/rest. Your actual pattern says Tuesday/Thursday/Saturday with occasional Sunday additions. A system that detects your real schedule beats one that assumes the plan.
Monotony and Strain: The Overtraining Signal Wearables Miss
ACWR catches load ramps but misses the shape of the training week. Monotony and strain fill the gap by flagging weeks without enough hard/easy variance.
What Is RED-S? A Literacy Guide for Athletes and Active Adults
Relative Energy Deficiency in Sport (RED-S) explained in plain language — the body systems it affects, who is at risk, what diagnosis actually involves, and why wearables cannot detect it.
Personalized Thresholds: Why p25/p50/p75 of Your Own Data Beats Population Norms
A resting heart rate of 68 is normal in the population and elevated for you. Personalized thresholds built from your own p25/p50/p75 catch meaningful shifts that population norms miss entirely.
IARC Group 1 and 2A Carcinogens in Food: Weekly Exposure, Not Single Meals
IARC classifies substances by strength of evidence, not risk magnitude. Processed meat, alcohol, and charred meat compounds are worth tracking weekly — as awareness signals, not alarms.
EWMA vs Rolling ACWR: Why the Week-Boundary Math Lies
Rolling-average ACWR creates week-boundary artifacts. EWMA ACWR smooths the signal, handles missed sessions, and produces trend information that matches physiology.
Energy Availability and RED-S: A Data-Literate Guide to Recognizing Underfueling
Energy availability, RED-S risk factors, biomarker signatures of underfueling, and the confounders that mimic them — plus where data stops and clinical assessment begins.
Why Your Muse Data Needs Cleanup: Artifact Detection for Consumer EEG
Consumer EEG is noisy. Muscle activity, blink, jaw clench, and electrode liftoff produce spectral signatures that drown the real brain signal. How autoencoder-based artifact detection separates brain from noise.
Your Diet Has a Pattern — Here Are the 8 Common Ones
Mediterranean, ketogenic, high-protein, plant-based, high-processed, balanced, low-carb, high-carb. Weekly logs reveal which pattern you actually follow — often different from what you think.
Your Bench Doesn't Progress Like Your Squat: Per-Movement Load Progression
Bench, squat, and deadlift progress at different rates for structural reasons. A one-size-fits-all progression rule under- or over-loads every lift.
When Oura and Garmin Disagree by 20+ Points: Cross-Source Validation
Two devices measuring the same night and reading 58 and 79 is telling you something — usually that the composite score shouldn't be a single number at all. How cross-source validation handles divergence.
Glycemic Index Per Meal (Not Per Food): How Mixed Eating Changes the Response
Per-food GI tables ignore that nobody eats a single food alone. Protein, fat, fiber, and cooking method all shift the curve. Meal-level glycemic load is closer to what your glucose actually does.
RPE Calibration: Why Your 9 Is Someone Else's 7 (and Why It Matters)
Self-reported RPE drifts with fatigue, stress, and experience. Calibration against objective markers (velocity, RIR, recovery) catches the drift.
Chrono-Nutrition: Eating Window, Meal Spacing, and Why Timing Changes Outcomes
When you eat changes outcomes independent of what you eat. Eating window, meal spacing, and circadian alignment are measurable, and they matter. Here's how to track them.
When Oura Syncs Late: How Source Fallback Chains Keep Scores Consistent
Your primary wearable didn't sync last night. Your dashboard shouldn't just go blank. Source fallback chains degrade gracefully across devices while preserving confidence.
Per-Muscle Recovery Half-Lives: Why Legs ≠ Shoulders ≠ Chest
Your chest isn't recovered in the same window as your quads. Here's how per-muscle recovery half-lives change training frequency and session design.
Polyphenol Diversity: Why Variety Beats Milligrams for Antioxidant Intake
Total polyphenols in milligrams misses the point. A diet hitting six polyphenol classes outperforms one hitting a single class at higher dose. Here's how class diversity tracking works.
Detraining: How Fast You Actually Lose Strength (and What the Research Shows)
Strength is more durable than aerobic fitness. Here's what the detraining literature shows about what you lose, when, and how to ramp back safely.
Readiness Score Confidence: When the Number Should Be Suppressed
A readiness score with no confidence value is an opinion, not a measurement. How to read credible intervals, spot low-confidence days, and know when to suppress.
NOVA Groups: Why 'Ultra-Processed' Isn't the Same as 'High-Calorie'
NOVA sorts food by processing, not calories. A 400-calorie bowl of oats and a 400-calorie bowl of sugary cereal differ in ways macros cannot capture. Here's how the four groups work.
Maximum Adaptive Volume (MAV): Why Per-Muscle Training Ceilings Beat Generic Volume Rules
MEV and MRV are population ranges. MAV is the per-muscle weekly ceiling your body can currently tolerate — learned from your history, not a table.
Nutrition Intelligence: What's Actually In Your Food Beyond Calories and Macros
NOVA processing, polyphenol diversity, meal-level glycemic index, chrono-nutrition, IARC exposure, and 35 micronutrients — the nutritional dimensions calorie counters ignore.
Adaptive Training Intelligence: The Load Signal Your Wearable Isn't Showing You
How per-muscle volume tolerance, recovery half-lives, and Bayesian load models translate raw wearable data into actionable training prescriptions.
When to Trust Your Health Score: Confidence, Cross-Validation, and the Limits of Wearable Data
Composite health scores fuse many inputs into one number — but only if you know which inputs are trustworthy. Confidence, cross-validation, suppression.
Body Composition: DEXA vs Smart Scales vs Calipers
Your bathroom scale's body fat number is a guess with a ±5% error margin. Here's what DEXA, bioimpedance, and calipers actually measure, how accurate each is, and how to use them together.
How to Interpret Your Sleep Score Across Devices
A 75 on Oura is not a 75 on Garmin. Here's what each sleep score actually measures, which factors predict how you feel, and why consistency matters more than any single number.
What Is a Composite Health Score and Why Does It Matter?
Single metrics lie by omission. A composite score synthesizes HRV, sleep, training load, and recovery into one number — but only if you can see how it's built.
What Is ACWR and Why Does It Matter for Training?
The acute-to-chronic workload ratio is the single best predictor of training-related injury. Here's what it measures, where the 0.8-1.3 sweet spot comes from, how Omnio calculates yours, and the mistakes that get people hurt.
What Your Blood Work Actually Tells You About Fitness
Lab reference ranges are built for the general population. If you train seriously, most of your results need a completely different interpretation.
How Wearables Measure Stress and Strain — and What the Numbers Mean
Garmin Body Battery, WHOOP Strain, Oura Stress, WHOOP Recovery — these scores sound like they measure the same thing. They don't. Here's what each one actually calculates, why they contradict each other, and how to read them together.
What Is HRV and How Do Wearables Measure It?
RMSSD, SDNN, LnRMSSD — your wearable picks one and doesn't tell you which. Here's what HRV actually measures, how Oura, Garmin, WHOOP, and Polar each capture it differently, and why your baseline matters more than any single number.
Best Bevel Health Alternatives for Deep Health Intelligence
Bevel is a polished health dashboard with Oura and Garmin support. But if you want validated nutrition data, transparent scoring, or cross-domain correlations backed by statistics — here's what else is out there.
Best Cronometer Alternatives for Nutrition Tracking
Cronometer is the gold standard for micronutrient tracking. But if you want your nutrition data connected to sleep, HRV, and training — or a modern mobile experience — here are the best alternatives.
Best Levels Alternatives for Metabolic Health Tracking
Levels made continuous glucose monitoring mainstream. But CGM data alone doesn't tell the whole story. Here are the best alternatives for understanding your metabolic health.
Best MacroFactor Alternatives for Adaptive Nutrition
MacroFactor's adaptive algorithm is best-in-class for macro coaching. But if you want micronutrients, meal quality, or wearable-connected insights, here are the alternatives worth considering.
Best MyFitnessPal Alternatives That Understand Your Diet
MyFitnessPal counts calories. These alternatives track meal quality, micronutrients, and how your diet affects sleep, recovery, and training. Here's what to switch to.
Predicting Health Dips Before They Happen
Your wearable data contains early warning signals that precede readiness dips by 1-5 days. We built a system that learns which signals matter for you personally — and warns you before the dip arrives.
Your Body Already Knows When to Focus
Your wearable data contains a hidden signal: the ultradian rhythm. We're building a system that reads it — predicting your best focus windows from sleep data, then refining in real-time with a heart rate monitor.
Most Nutrition Trackers Count Calories. Ours Understands Your Diet.
Calorie counting is table stakes. Omnio validates your logs against government nutrition databases, scores meal quality, tracks 35 micronutrients and polyphenols, classifies your dietary pattern, and connects what you eat to how you sleep, recover, and train.
Why Connecting Accounts with OAuth 2 Doesn't Share Your Password
When you connect Oura, Whoop, or Withings to Omnio, you sign in on their website, not ours. Here's what OAuth 2 does, what we receive, and what we never see.
Your Calorie Target Is Wrong. Here's How We Fix It.
Most calorie calculators give you a number based on a formula from the 1990s and call it a day. Omnio watches what you actually eat, how your body actually responds, and learns your real metabolic fingerprint over time.
Your AI Health Assistant Doesn't Know Who You Are
Your AI health assistant analyzes your sleep, recovery, and training data — but it never sees your name, email, or account ID. Here's how we built privacy into the architecture itself.
Can Your Health AI Prove It Isn't Lying to You?
We asked Oura, WHOOP, and Omnio the same research question. Two gave vague platitudes with zero citations. One gave verifiable evidence. Here's why the difference is architectural.
Adaptive Training With Wearable Readiness Data
Omnio doesn't just track your workouts — it derives your training schedule, gates every session through biometric readiness, adjusts for nutrition, and learns from outcomes. Here's how we're building a genuinely adaptive training system.
Which Wearable Is Most Accurate? What 17 Validation Studies Actually Found
We reviewed 17 peer-reviewed studies comparing Oura, Apple Watch, Fitbit, Garmin, WHOOP, and Samsung across sleep, heart rate, HRV, SpO2, VO2 max, and more. Here are the specific numbers.
Why We Built a Bayesian Brain for Your Training Plan
Every fitness app says it 'learns.' We wanted to prove it. Here's why we chose Bayesian parameter estimation over neural nets, how six independent sub-models personalize your training, and why the system can never be worse than a textbook.
Your Wearable Says Rest — But Should You Actually Train?
Readiness scores tell you how recovered you are. They don't tell you whether skipping today's session is smart load management or the start of detraining. We added a sports science metric to close that gap.
How We Avoid Spurious Correlations in Health Data
When you test enough metric pairs, something will appear correlated by chance. Here's how Omnio's correlation engine uses detrending, dual methods, p-values, and FDR correction to separate real signal from noise.
We Asked Oura, WHOOP, and Omnio the Same Sleep Question. Here's What Happened.
Every major wearable now has an AI assistant. We asked all three to compare a week of sleep data. The difference in depth reveals a fundamental architectural gap.
Why We Built a Model Context Protocol Server for Health Data
Most health platforms treat AI as a chatbot skin over a database. We built an MCP server with a plugin architecture, time-series storage, and multi-step reasoning. Here's why.
How to Combine Oura and Garmin Data in One Dashboard
Using Oura for sleep and Garmin for training? Here's how to combine both data sources into one dashboard without duplicating metrics or losing context.
Garmin vs Oura for Recovery, Readiness, and Sleep
Garmin and Oura solve different problems. Here's how they compare for sleep tracking, training readiness, workouts, HRV, and whether you should use one or both.
Oura vs WHOOP for Sleep, Recovery, and Training
Trying to choose between Oura and WHOOP? Here's the practical difference in sleep tracking, HRV, recovery, workouts, battery life, and who each device is actually best for.
Best App for Polar H10 Heart Rate and HRV Tracking
Looking for the best app for a Polar H10 chest strap? Here's what to look for if you care about heart rate, RR intervals, HRV, ECG, guided readiness checks, and long-term trends.