How Biohacking Kompakt works

How the research is done

Research starts with the primary literature, mainly via PubMed and Europe PMC, plus Cochrane, clinicaltrials.gov and the statements of EFSA, BfR, EMA, DGE and WHO. The order of evidential weight is fixed: meta-analyses and systematic reviews, then randomized controlled trials, then cohorts, case series, animal models and cell culture.

Not sources: supplement shops, manufacturer blogs, peptide vendors and content that itself only summarizes. A manufacturer’s claim is cited only if it is explicitly labelled as such.

How numbers are checked

Research and text drafts are produced with AI support. For every topic page there is an evidence file in which every number in the text is listed with its source. Before publication, an automated check compares every number in the finished text against this evidence file. Whatever is missing there is removed from the text. No calculations are made: if a study says “week 36” and “week 88”, both appear on the page, not the difference.

Whatever cannot be verified is left out. Instead, the section “Where the data stop” states that there are no reliable figures on it.

The BK-Score: five axes from 0 to 10

Each axis rates what is known. None rates whether something is good. The values are a disclosed assessment by Biohacking Kompakt according to the following rules, not a scientific evaluation and not a recommendation. The complete dataset is available as scores.json.

Human evidence

How much data from humans exists and how good it is. 0 to 1: no human data. 2 to 3: case reports and uncontrolled series. 4 to 5: small or unblinded studies, often only with laboratory values as the endpoint. 6 to 7: several randomized trials, but small, short or inconsistent. 8 to 9: large randomized trials or consistent meta-analyses with hard endpoints. 10: approval for exactly this use.

Mechanism

How well it is understood why something is supposed to work. 0 to 2: a marketing narrative. 3 to 5: plausible from cell culture and animal experiments. 6 to 8: target structure confirmed in humans. 9 to 10: chain of action measured in humans.

Safety data

How well safety has been studied, not how safe something is. 0 to 2: not systematically studied. 3 to 5: short-term data only. 6 to 8: controlled data over months. 9 to 10: long-term data from approval and pharmacovigilance. A high number means “well studied”, not “harmless”.

Hype gap

How far the marketing runs ahead of the data. 0 to 2: promises without a basis. 3 to 5: a single finding is turned into a general claim. 6 to 8: mostly marketed accurately. 9 to 10: advertising in line with the data.

Track record of use

How much documented use in humans exists. Explicitly does not measure whether something works. 0 to 1: practically none. 2 to 3: only in studies. 4 to 5: widespread, but without a regulatory framework. 6 to 7: widely used for many years and regulated in one country. 8 to 9: decades under supervision. 10: decades in millions of people.

The label

No average is formed from the axes. What matters is the direction of the data: high human evidence only means that a lot has been studied, and studied well. If the good studies come out negative, the label is not “well supported” but “effect not confirmed”.

• Well supported – human evidence 8 or higher, direction positive or mixed, advertising in line with the data.
• Well supported, heavily overhyped – human evidence 8 or higher, but the marketing promises considerably more.
• Supported, with caveats – human evidence 6 to 7.
• Thin human evidence – human evidence 4 to 5.
• Long used, barely studied – few studies, but a lot of documented use.
• Well studied – effect not confirmed – a lot of good data, and it speaks against the advertised benefit.
• Studied – no benefit shown – some data, and it shows no benefit.
• Not studied in humans – there are no human data.
• Hype far ahead of evidence / Experimental – new, barely studied, in some cases heavily advertised.

Legal limits

For substances that are not approved, the site gives no dosage. For approved substances and nutrients, it states what the studies used and which reference values and upper limits authorities recommend. That is a statement of fact about studies, not an instruction for use.

Updates and corrections

Every page states its last-updated date. When new studies change the assessment, text and score are adjusted; larger changes are listed in the change log (German). Please report errors, ideally with a source, to kontakt@biohackingkompakt.de.