Every longevity compound Luna has indexed.

A compound deep-dive at Heaven Health is one canonical page per molecule — its mechanism in plain English, an honest read of the human evidence, the doses studied in research, the risks worth knowing, and when Luna would or would not bring it up inside the app. New compounds enter the library as the research engine surfaces them. There are no affiliate links and there is nothing to buy here.

What a deep-dive looks like

When you open any individual page under /compounds — for example rapamycin, spermidine, or NMN — you will find the same structure. Quick facts at the top: drug class, mechanism, regulatory status, evidence rating. A plain-English explanation of how the compound works at the cellular level. A table of the human and animal studies the research engine has read, each row linking back to the underlying paper analysis on /research. The dose ranges that have been studied — framed as what the literature did, not what you should do. The risks and contraindications, prominent and honest. And finally: what Luna does with this compound. Whether she would consider it as part of a protocol, under what conditions, and why or why not.

That last section is the one most other sites do not have. Wellness blogs tell you what a compound is. Supplement vendors tell you to buy it. We tell you the third thing: how the assistant you might actually use makes a decision about it. That is the thing that matters once you have read enough.

The classes of compound this hub covers

Every compound in the library belongs to one of the classes below. The classes themselves are a useful map of the longevity field. If you are new to this, read down the list — it is a reasonable five-minute summary of what humanity is currently trying.

How a compound earns a page

The research engine sees a compound for the first time when it appears in an analysed paper. The compound name is extracted into the canonical compounds collection. From there, several thresholds gate publication. A compound gets a published deep-dive only when the engine has analysed at least a small set of independent papers; when at least one human study exists (or when the absence of human data is itself the relevant story, in which case the page is explicit about that); and when the medical reviewer has signed off on the mechanism description, the evidence summary and the risk section.

That is why this hub is not a comprehensive supplement encyclopaedia. The library is growing, deliberately, with care. Compounds in active drafting include several senolytic combinations under trial, two newer mitochondrial peptides, and the cluster of GLP-1 follow-on agents. They will appear here when they are ready.

The framing we try to hold

There are two ways to write about longevity compounds. The first is to make every page sound like a press release: "Compound X has been shown to..." with no mention of effect size, study design, or whether the finding has held up in a second laboratory. The second is to write it the way you would want a friend who happens to be in clinical research to write it: this is what the evidence shows, this is what we do not know, this is where the field is honest about its uncertainty, this is what would change my mind.

Heaven Health publishes the second kind. Sometimes that means a compound you have heard a lot about has a deep-dive that ends with "the evidence does not yet support this and here is why we are watching it anyway." That is the version that earns the right to be trusted.

The best compound page is one that survives the next ten years of research without needing to be quietly deleted.

What this hub is not

A short list, because it matters.

• Not a shop. No affiliate links, no buy buttons, no sponsored placements. The compounds we cover are not commercial relationships.
• Not a prescription guide. We do not publish dosing instructions for prescription medications. The page tells you what doses appear in the literature; what to do about that is a conversation with a clinician.
• Not a replacement for clinical advice. Anything you read here is educational. Heaven Health is built around an AI assistant who is informed and patient — not around an AI that should be making medical decisions for you.
• Not finished. The library will keep growing as long as the research engine keeps running. Some categories above already have multiple pages; some have one; some are placeholder anchors for what is being drafted.

How to read a compound page well

If you have not spent a lot of time in this literature before, a short guide. When you open a deep-dive, start at the bottom — the risks and contraindications section. Knowing what could go wrong is the right frame for everything above it. Then read the mechanism, slowly. Most longevity compounds work by adjusting one of a small number of pathways — mTOR signalling, NAD+ availability, autophagy, sirtuin activity, mitochondrial function, inflammation. Once you have read three or four mechanisms you will notice the pathways repeat. That repetition is the field's actual structure.

Then read the evidence section, but read it for effect size as much as for whether the finding exists. A compound that nudges a biomarker by one percent in a small study is not the same as one that moves it by twenty percent in a large trial. The page is built to make that distinction visible. Finally, read the "what Luna does" section. This is where the page commits to a position. We try to be clear about it. Some compounds Luna will routinely raise when relevant; others she mentions only when a user specifically asks; others she actively does not recommend because the evidence does not yet justify the risk profile.

A few honest patterns the library reveals

Reading across many compounds for long enough surfaces patterns the field does not always advertise. Three worth knowing about.

Most compelling animal evidence does not translate to humans. The literature on senolytics, NAD+ precursors, sirtuin activators and several other classes is striking in mice. The human evidence, where it exists, is more modest. That gap is not a scandal — it is the normal pattern of translational biology. But it is worth holding in mind every time you read "studies show" on the internet.

Combination signal often beats single-compound signal. Many of the strongest patterns in the engine's reading are not "Compound X moved Biomarker Y" but "Compound X plus Behaviour Z plus better sleep moved Biomarker Y meaningfully." That is one reason the protocols layer exists. Compounds alone rarely tell the whole story.

The plainest interventions usually have the strongest evidence. The compound pages with the heaviest citation lists are rarely the exotic ones. They are vitamin D, omega-3, creatine, magnesium. None of them sound exciting. All of them have decades of well-funded research. Most of the time, doing the basics correctly is the unfinished business — not adding a peptide on top.

Adjacent hubs

• Research — the per-paper analyses that each compound page draws on.
• Protocols — structured plans that combine specific compounds with behaviours and tracking.
• Blog — long-form editorial pieces that connect multiple compounds across one theme.
• Longevity — the pillar page tying the whole landscape together.
• Luna — meet the assistant who uses all of this in your daily plan.
• Heaven Score — how compounds and behaviours feed into one daily number.
• Biological age — the biomarker side of the same story.