Source policy

Nutrition content is only as honest as its sources. This policy governs what GLP-1 Diet Tab is allowed to cite, how sources are tracked, and what happens when they age.

The hierarchy we use

We prefer, in order: peer-reviewed research and systematic reviews; guidance from bodies such as the FDA, NIH, CDC, and the Academy of Nutrition and Dietetics; official prescribing information for context on how GLP-1 medications affect appetite; and named-expert clinical commentary. Consumer blogs, supplement marketing, social media claims, and press releases are not sources here, no matter how popular. If a claim cannot be supported from the acceptable tiers, we soften it or cut it.

Dated, tracked, and revisited

Every source in our editorial system carries the date we verified it and a freshness window. When the window lapses, the source is automatically flagged stale and the content that depends on it returns to review. Nutrition science moves; a 2021 citation about appetite effects may deserve retirement in 2026, and our system makes that visible instead of silent.

How we handle disagreement

Where credible sources disagree, we say so in the text rather than picking the convenient answer. Where evidence is thin, especially around newer peptide therapies and food interactions, we label the uncertainty explicitly. Members make better decisions with honest error bars than with confident guesses.

Corrections

If you believe a source is misused, outdated, or missing, tell us through the contact page. Reports go to a human reviewer, corrections ship promptly, and substantive changes are noted on the affected page. We would rather fix a citation today than defend one tomorrow.

Concrete examples

Acceptable: a peer-reviewed satiety study in a recognized journal, FDA prescribing information used to describe appetite effects in general terms, or USDA nutrient data behind a recipe’s protein figure. Not acceptable: a supplement brand’s blog post about peptides, a viral social media claim about food timing, an affiliate review site with no methodology, or a press release repackaged as research. When members send us the second kind, we explain why it cannot support a claim here.

AI and sources

Our AI drafting pipeline is not allowed to introduce sources on its own authority. Claims in AI-assisted drafts must trace to material already vetted into our source system, and reviewers specifically check for invented or misattributed citations, a known failure mode of language models, before anything publishes. A citation that cannot be located and verified by a human is removed along with the sentence leaning on it. The result is slower publishing than a content farm and considerably fewer regrets, which is exactly the trade a food platform serving people on prescription medication should make.