The Vitality Score is a 0–100 wellness indicator calculated from Apple Health data that a user permits Health Genie to read. It is designed to summarize recent movement and recovery-related signals. It is not a clinical score.
Inputs and baselines
| Input | Comparison baseline | Fallback when history is unavailable |
|---|---|---|
| Steps today | Rolling 7-day average | 7,500 steps |
| Resting heart rate recorded for today | Average of the available resting-heart-rate history loaded for up to 60 days | 72 bpm is used for today’s input when no current-day value exists, and for the baseline when no history exists |
| Last night’s sleep, when available | Rolling 7-day average | 7.5 hours |
The fallback values are product defaults used so the app can produce an initial result before enough personal history exists. They are not medical targets or diagnoses.
Formula without sleep
When no positive sleep-duration value is available, the score uses steps and resting heart rate:
heart-rate trend = (resting-heart-rate baseline ÷ today’s resting heart rate) × 100
raw score = (step trend × 0.55) + (heart-rate trend × 0.45)
Formula with sleep
When a positive sleep-duration value is available, the weights change and sleep is added:
raw score = (step trend × 0.45) + (heart-rate trend × 0.35) + (sleep trend × 0.20)
The final score is rounded to the nearest whole number and limited to the range 0–100. Capping step and sleep trends limits how much a single unusually high value can inflate the result.
Score labels
| Score | Label |
|---|---|
| 70–100 | Great |
| 40–69 | Fair |
| 0–39 | Low |
These labels describe the product’s own score bands. They are not clinical categories and should not be interpreted as proof that someone is healthy or unwell.
What the score does not use
This version of the Vitality Score does not use HRV, blood oxygen, ECG results, wrist temperature, medications or medical history. AI does not calculate the score; the deterministic formula above does. AI may be used separately to turn selected metrics into a plain-language brief.
Known limitations
- Missing or inaccurate sensor data can affect the result.
- Population fallback values make early scores less personalized.
- A lower resting heart rate is not always better; individual context matters.
- A single score cannot account for illness, medication, pregnancy, disability or other personal circumstances.
- The score has not been presented as a clinically validated diagnostic measure.