Longevity protocols, designed from the research.

A protocol at Heaven Health is a structured plan — usually 30, 60 or 90 days — that combines specific compounds, specific behaviours and a specific set of biomarkers to track. Each one is designed by Luna from peer-reviewed evidence in the research engine, signed off by a medical reviewer, and written in plain English. You can read them here. You can run them inside the app. You should not treat any of them as a prescription.

What "protocol" means at Heaven Health

The word gets used loosely in the longevity world. Sometimes "protocol" means a celebrity's morning routine. Sometimes it means a single supplement. Sometimes it means an aspirational chart pinned to the wall of a biohacker's garage.

At Heaven Health it means one specific thing: a structured plan with four elements. Compounds — what to consider taking, at what timing, with what evidence. Behaviours — what to do, when, for how long. Tracking — which biomarkers to actually measure to know whether the plan is doing what it claims. Time-bound — every protocol has a start and an end, because the discipline of a finish line is what turns a plan from a vibe into evidence.

To date the research pipeline has generated 82 of these. They are not all live as public pages yet. The library publishes a protocol only after the medical reviewer has read every step, checked the underlying citations on /research, and confirmed the safety section is honest. That filter is what makes the protocol layer different from a wellness blog. You should expect new protocols to appear here at roughly one per week as the queue clears.

The categories of protocol this hub covers

Every protocol lives under one of the categories below. The categories themselves are a useful tour of where current longevity research has its strongest signal. Each category links to the canonical protocol page where one exists, or to the research feed where it does not yet.

How a protocol is built

Each protocol page begins life inside the research engine. The protocol-designer agent looks at the strongest signals in the hypothesis layer and asks: if this is real, what would a reasonable plan look like? It drafts a structure — a daily and weekly cadence, a supplement list, a behaviour list, a tracking list. Every step is tagged with the underlying citations.

What we will and will not put in a protocol

Some plain rules, because the topic deserves them.

We will name compounds by their generic name. Describe the doses that have been used in published research. Name the biomarkers worth tracking. Be specific about timing, ordering and combinations. State the expected effect size and admit when it is small. Say when the evidence is good and when it is preliminary.

We will not tell you what to take. Specifically, no protocol on this site contains a "you should take X mg per day" instruction for any prescription compound. The protocol will tell you what doses appeared in the underlying studies. What to do about that information is a conversation between you and a clinician. That line is non-negotiable.

This is the slow path. It is also the only path that earns durable trust on a subject as important as your own body.

A protocol is only useful if you can finish it and know whether it worked.

How protocols feed the rest of the app

Reading a protocol on this site is one thing. Running one inside the app is another. When you open Luna and pick a protocol, three things happen automatically. Your Heaven Score dashboard adds the protocol's biomarkers to your tracking. Your daily plan adapts: morning ritual, meals and movement adjust to the protocol's cadence. And Luna knows the protocol in detail — you can ask her any question about any step, and she will answer from the same research the protocol was built from.

The protocol layer is also where the biological age story becomes practical. Most protocols are framed around measurable change in a specific biomarker over a specific window. That is exactly what a biological age clock is built to capture. Protocols are how you turn the research into something an epigenetic readout can score.

Categories of protocol coming next

The 82 designed protocols include several that are still being reviewed. Categories the queue includes:

• Cardiovascular longevity — ApoB-led plans with lipid-management and dietary architecture.
• Cognitive resilience — sleep, exercise and nutritional levers that map to cognitive ageing markers.
• Hormonal alignment — women's protocols around perimenopause, men's around mid-life androgen drift.
• Recovery-from-illness protocols — what the research suggests in the weeks following a significant infection.
• Skin biology — the small but real intersection of dermatology and the hallmarks of aging.

How a protocol page is structured, in practice

Every protocol page on this site follows the same opening shape, so you can scan it in thirty seconds before committing the time to read it properly. At the top: a one-line description of the goal. A duration. An evidence rating — strong, moderate, preliminary, or animal-only-so-far. A short list of who this protocol is and is not for. Then the structure of the plan: daily cadence, weekly cadence, milestone points. Then the supplement list with timing. Then the behaviour list. Then the biomarkers worth tracking. Then the citations.

This is deliberate. A protocol you cannot read at a glance is one you will not actually follow. The structure is also designed to be the same shape whether the protocol is a tightly defined 30-day senolytic pulse or a year-long circadian-alignment plan. Same skeleton, different content. After three or four protocols the format becomes second nature.

What makes a Heaven Health protocol different

There are now several places on the internet to find a "longevity protocol." Some are charismatic morning routines from public figures. Some are dense supplement stacks from biohacker forums. Some are clinical interventions that require a physician to even discuss. Where does this fit?

Three things distinguish a Heaven Health protocol. First, it is built from primary research, not from a personality. The starting material is the same literature read by working geroscientists — not what one person decided was working for them this quarter. Second, it has a finish line. Every protocol commits to a duration and a measurable outcome. If you cannot tell at the end whether it worked, the protocol was not actually a protocol. Third, it lives next to an assistant who knows it. Reading a protocol on a static page is useful. Running one inside the app — with reminders, daily check-ins, biomarker logging, and Luna's working knowledge of the underlying citations — is the version that earns its keep.

What we learn from the protocols that did not get published

The pipeline has generated 82 protocols. Not all of them will become public pages. Some get parked in review because the underlying evidence is not yet strong enough — interesting, but not ready. A few get parked because the safety profile of one of the compounds means we are not willing to write a public plan around it without a more developed clinical context. A few get redrafted because the medical reviewer flagged a step we cannot make precise enough without crossing the line into prescription territory.

That filtering is part of the value. The protocols that do make it through are protocols we are willing to put a name on and a date on. A short library of carefully-built, reviewed plans is more useful than a long library where most entries you would not actually recommend to a friend. Quality compounds over time. Volume does not.

Adjacent hubs

• Research — the per-paper analyses each protocol draws on.
• Compounds — every molecule referenced inside a protocol, on its own canonical page.
• Blog — long-form editorial pieces that connect protocols across a single theme.
• Longevity — the pillar page that ties the whole landscape together.
• Luna — the AI assistant who runs the protocols with you.
• Heaven Score — the daily metric every protocol nudges.
• Biological age — the long-arc outcome most protocols are trying to move.