The version of a nutrition study that reaches you is usually about 60 words, and it carries none of the three facts that decide whether it applies to you: the dose, the duration, and who was in it. Those three live in the record, and they're the only reason we link to one.
What a record has that a story doesn't
Take a trial that comes up constantly in circulation research. Sixty-five firefighters took 1,200 mg of aged garlic extract with 120 mg of CoQ10 every day for 12 months, and the team measured arterial stiffness and how well blood vessels responded. Against placebo, pulse wave velocity fell 1.21 m/s. Every number in that sentence is load bearing, and a write-up that keeps the result and drops the 12 months has changed the meaning without changing a word of the finding.
The record also tells you the design. Randomized, double blind, placebo controlled means one thing. An open label pilot in 18 people means another, and both get described as a study in the press.
Figures drift as they travel, too. A relative change gets written up as an absolute one, a subgroup result becomes the headline result, and a finding in people who started with elevated readings becomes a finding in everyone. The methods text is where the limits are written down, usually by the people who understand them best, and it sits three paragraphs below the part anyone quotes.
The three lines people skip
Sample size comes first. Sixty-five people is a real trial and a small one, and small trials are noisy enough that a single unusual participant shifts the average. Duration comes second, because a nutrient that changes something over 12 weeks may show nothing at 2 weeks, and a lot of disappointment comes from judging on the wrong clock. Dose comes third, and it's the one that decides whether the finding has anything to do with the bottle in your hand. A trial run at 1,200 mg tells you nothing about a 300 mg serving.
Endpoints deserve the same treatment. A lot of work in nutrition measures a marker rather than something you'd notice, and a change in a marker is a finding about the marker. That doesn't make it worthless. It does mean the accurate sentence is longer than the version that gets shared, and if a page can't tell you what was actually measured, it hasn't read the paper.
None of this requires training. It requires scrolling past the conclusion.
Funding changes the number, quietly
A fair share of the larger aged garlic trials list the extract manufacturer as a funder, and pooled analyses written by investigators with that funding tend to report larger effects than independently funded ones. Industry money doesn't make a result fake. It does shift where the emphasis lands, which outcomes get reported, and how the abstract is worded, and you'd never see any of that from a summary.
The same caution applies to us. When we describe a trial, we name the dose, the number of participants and the endpoint, and we say that the work was done on the ingredient rather than on the finished product, because no trial of the finished product exists. A claim about a nutrient in a published paper is not a claim about what's in your cupboard, and the distance between those two sentences is where most supplement marketing lives.
How we handle it here
Every citation on this site points at the primary source, which is the journal record or its PubMed entry, so you can read the abstract yourself and check whether we described it fairly. We link to PubMed rather than to a news page because a news page can be edited, taken down, or written by someone who never opened the paper. The record stays put.
One piece of housekeeping worth knowing. Under the 1994 law that governs this category, a structure or function statement has to be notified to the FDA within 30 days of first marketing, and the disclaimer you see under our descriptions is part of that requirement. It isn't boilerplate we chose. If you take prescription medication or you're managing a condition, take the abstract to your provider rather than to us, and give any change 4 to 8 weeks of consistent use before you call it a result.
Read the source when you can. It takes four minutes and it beats trusting anyone's summary, including ours.
Frequently asked questions
Where do your citations go?Show answerHide answer
To the primary source: the journal record or PubMed entry.
Why not link to a summary article?Show answerHide answer
Because summaries drop the details that matter, especially dose and population.
Do you cite studies of the finished product?Show answerHide answer
No. Studies are on the ingredients. We say so explicitly.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
