Research

The protocol is assembled from published research on lifestyle interventions for blood pressure. Every lever in it traces to a meta-analysis or trial we cite by name. What we do not do is collect data back from the people using it.

Where the protocol comes from

Each intervention has its own page listing the evidence tier, the effect size reported in the literature, and a direct link to the source. The largest single input is the Edwards et al. 2023 meta-analysis of isometric exercise — 270 trials, n=15,827 — which is why wall sits sit at the front of week one rather than a step-count target.

Read the evidence base at /science →

What we collect from users

Nothing. There is no account, no server and no analytics SDK in the app, so there is no dataset of user outcomes — not an anonymised one, not an aggregate one, not one held briefly and discarded. Your readings and your protocol history stay in storage on your iPhone.

That rules out a category of thing we would genuinely like to build. A large opt-in outcome set would let us tell you how your wall-sit result compares to other people who train like you, instead of to a trial population. We have not built it, and we will not describe it as though we had. If that ever changes it will be opt-in, it will be announced here first, and this page will say so.

What the app does instead

The correlation work happens entirely on your device, against your own record. The lever view compares the readings you logged on days you kept a habit against the days you did not, and reports the difference. It is a description of your own data — a sample size of one, with all the caveats that carries. It does not predict an outcome and it is not a clinical finding.

How that is built, in detail, at /privacy-engineering →