The agency mailed it on a Thursday. Four pages, single-spaced, addressed to a North Carolina firm operating out of a strip-mall storefront in Fayetteville. Cape Fear Naturals had been compounding small-batch supplements and selling them through a regional retail counter and a serviceable little e-commerce site. Nothing in the formulation was unusual. The Warning Letter was not about the products. It was about the language.
Twenty-one separate claims were quoted in the letter, drawn from the company website and product labels. The FDA central allegation was familiar to anyone who has read a CFSAN letter: by claiming that dietary supplements diagnose, prevent, treat, cure, or mitigate disease, Cape Fear Naturals had effectively introduced the products as unapproved new drugs.
What is interesting about this letter is the granularity. The FDA did not list one or two flagship claims. It listed twenty-one. Each was a precise phrase, lifted intact from the brand marketing copy.
What the letter actually quoted
Two stand out as obviously past the line:
Prevent and treat diarrhea. Relieve irritable Bowel Syndrome, Crohn disease, and gut dysbiosis. Decrease frequency of vaginal yeast infections and urinary tract infections. Treat respiratory infections. Improve Blood Pressure and Cholesterol.
Five named diseases in fifty words. A reasonable reader could not mistake that paragraph for a structure-function claim under 21 CFR 101.93(g). The letter cited it whole.
Why this letter still matters
Cape Fear Naturals was a small operation. The civil action did not become a class action. The brand quietly revised its copy and the letter dropped off the trade press radar within a week. By any operational measure, it was routine enforcement.
But routine enforcement actions are precisely what AI engines now consume when answering questions about supplements. We index every Warning Letter back to 2017. We embed every quoted phrase. Cape Fear twenty-one phrases — once searchable only inside the FDA own document index — now reach the engines.
Three implications follow. First, the disease-claim line is precise but invisible. Second, the engines are getting their training signal from this corpus too. Third, small brands set precedent for big brands.