Methodology

Research sources

Where our evidence comes from, how we weight it, and what we mean when we say a pair is backed by research.

The bar we set

Human research, published, peer-reviewed

Every pair in our library requires published human research demonstrating a meaningful interaction between the two nutrients. Mechanism alone is not enough. Animal data alone is not enough. Anecdote, influencer endorsement, and marketing copy are ignored.

When a clinical trial exists, we cite it. When a meta-analysis synthesizes multiple trials, we use the meta-analysis. When evidence is thinner (mechanistic work plus observational data) we still include the pair only if the theoretical basis is robust, and we say so explicitly in the evidence rating.

Where we look

Primary databases and journals

The bulk of our research is sourced from the following databases and journals. These are the standard tools a working nutrition researcher would use, and they are all publicly accessible (abstracts at minimum, full text in many cases).

Journals we read

Peer-reviewed publications

Studies cited in our deep dives typically appear in these journals. We do not limit ourselves to this list, but it reflects the majority of the evidence base:

How we weight evidence

Our evidence tiers

Not all evidence is equal. When we rate the strength of a pair, we use a simple tiered framework. The evidence rating appears alongside every pair in the library.

Tier 1
High confidence. Multiple randomized controlled trials in humans, ideally supported by a meta-analysis or systematic review, demonstrating the synergistic effect. Dose ranges and timing are well characterized.
Tier 2
Moderate confidence. At least one well-designed human RCT plus supporting mechanistic or observational data. Dose-response is reasonably well understood but some parameters remain open.
Tier 3
Emerging evidence. Strong mechanistic basis with limited but positive human data, typically small trials or pilot studies. Included because the biology is sound, but we flag it clearly.
Tier 4
Theoretical only. We do not publish pairs at this tier. If a synergy is compelling in theory but has no published human data, it stays off the site until that changes.
What we exclude

Evidence we do not treat as evidence

Industry-funded studies with undisclosed conflicts. We flag funding sources and discount studies where the funder has a direct financial stake in the result.

Single-arm case reports. Interesting but not sufficient to ground a recommendation.

Animal studies as primary support. Useful for mechanism, but dose translation and species differences mean they cannot stand alone.

Blog posts, influencer claims, and marketing white papers. If a claim only appears in promotional material, it does not count as evidence.

Evidence evolves. We update. New trials get published, meta-analyses get revised, and consensus shifts. We review the library periodically and adjust pair entries and evidence tiers as needed. If you spot a study we should be citing, tell us.

For how we translate evidence into product picks, see our affiliate disclosure. For the broader mission, see the about page.

Last updated: April 2026