Updated 2026-08-25

What Your Heart Rate Looks Like Next to Your Calendar

I used to think a heart rate spike meant exercise. Then I lined mine up against my calendar and found the spikes had names.

Apple Health has been collecting heart rate data for years. Most people open it, see a number, think “that seems fine”, and close it again. The data is good. The question it answers is not very interesting.

The number is not the signal

A reading of 108 BPM tells you nothing on its own. It could be a flight of stairs, a difficult conversation, a cold, or a coffee. The number is identical in every case. What makes it meaningful is what you were doing at 2:04 PM.

That context already exists on the same phone, in the calendar. Nobody was joining the two.

What showed up immediately

The first week of looking at the two together produced things that were obvious in hindsight and invisible before.

Certain recurring meetings had a consistent, repeatable signature. Not “a stressful week” — a specific slot, every time it ran. That is actionable in a way that a weekly average never is.

The reverse was also true. Blocks that felt draining sometimes showed nothing at all, which is its own useful information: the feeling was real, but it was not physiological arousal, and the fix was probably not breathing exercises.

Why it stays on the phone

This is health data crossed with calendar data. Between them they describe your body and your movements, and that combination is about as sensitive as consumer data gets.

So there is no account and no upload. The app reads from Apple Health with permission, reads the calendar with permission, draws the two together locally, and that is the whole architecture. It is a smaller product for it, and much easier to reason about.

The design problem nobody warns you about

Showing two time series at once is easy. Showing them so a person can read the relationship in under three seconds is not.

We went through several versions that were technically complete and practically useless — everything visible, nothing legible. What ended up working was colouring the trace by whether the rate sits in or out of the normal range, and letting the calendar event label sit under the chart rather than on it. Less information on screen, more of it actually read.

One honest limitation

This is not a medical device and it does not try to be. It does not diagnose, it does not alert, and it will not tell you whether a pattern matters. It shows you your own two datasets next to each other and leaves the interpretation to you and, if it looks worth asking about, your doctor.