Our commentary
What USA TODAY reported, and what it got right
USA TODAY's Mary Walrath-Holdridge sets the International Food Guide Pyramid, published online on 23 September in the peer-reviewed journal Advances in Nutrition, beside January's US pyramid. As USA TODAY reports it, the international version rests on plant foods, healthier fat and protein sources and carbohydrate quality, while the US version moved grains to the smallest group and emphasised red meat and whole-fat dairy. The story keeps the caveats that matter: the US still caps saturated fat at 10% of calories, health groups welcomed parts of the US guidance, and the international pyramid doesn't exclude animal foods. It also quotes Karen Peterson of the University of Michigan, who says red meat and high-fat dairy advice without portion guidance undercuts that limit.
The headline disagreement is real. The paper and the US guidelines differ on protein (1.2–1.6 g/kg in the US, against the 0.8 g/kg the paper calls "established"), dairy (three daily servings including full-fat, against about one) and red meat, where the US sets no limit. Harvard's arithmetic supports Peterson: on 2,000 calories the cap is about 22 g of saturated fat, and three full-fat dairy servings use about 17 g. What the story doesn't tell readers is what kind of document the pyramid is, or how its agreement was built.
It is a proposal
The authors call it a "proposed" framework and "an evidence-informed, consensus-based narrative synthesis and conceptual framework". They say that doesn't mean "derived from newly conducted systematic reviews or formal evidence grading processes". Consensus came from author discussion "rather than formal voting or Delphi procedures", and "No new data were generated or analyzed." Nothing in the paper reports the pyramid being tested or adopted by any government or agency.
That changes how USA TODAY's "more than 250 scientific publications" should read. The number comes from BYU's press release, not the paper's text, and it counts what was cited rather than what a systematic search found. The "first global" framing also rests on that release. The paper doesn't claim it.
How the agreement was built
The US input wasn't the guidelines being compared. The paper's method lists "the 2025 United States DGAC Scientific Report". That is the advisory committee's report, which Harvard says the administration rejected in favour of a supplemental analysis. So the pyramid converges with the document January's guidelines departed from. The paper says this openly and treats the published guidelines as a divergence, so the comparison is legitimate, but "they analysed US guidelines" is loose.
The proportions came first. Step one chose the EAT–Lancet reference diet, whose intake ranges set the food-group proportions. National guidelines then gave "qualitative support", and the authors say their selection is "not intended to provide an exhaustive global representation". Co-author Walter Willett co-chairs the EAT-Lancet Commission. So the paper's figure of about 14 g of unprocessed red meat a day (15 g elsewhere in the paper) is the EAT–Lancet number, not a consensus of national guidelines. We haven't read the supplemental tables holding the actual guideline comparison.
The disclosures, on both sides
USA TODAY doesn't list them. The paper reports no funding, and its conflict statement runs beyond the meat and dairy interests the US guidelines are criticised for. Reported ties across several authors include Barilla, a glycemic index testing service, the Almond Board of California, the Soy Nutrition Institute, the United Soybean Board, Alberta Pulse Growers and the International Nut and Dried Fruit Council. The pyramid elevates low-GI staples, durum pasta, nuts, soy and legumes. Harvard, meanwhile, reports reviewers' disclosed financial ties to beef and dairy in the US supplemental analysis. Disclosure doesn't show bias on either side, and we infer none. Harvard isn't neutral ground either: Willett is a colleague of its authors.
Not simply strict versus lax
On some points the US is the stricter document. It says "no amount" of added sugar is recommended, with no more than 10 grams per meal, while the pyramid argues against treating added sugar as one category and singles out sugar-sweetened beverages. The US says to avoid highly processed foods, while the pyramid "does not rely on degree-of-processing classification systems".
The US text is also less one-sided than its graphic. It lists animal and plant proteins without ranking them and tells readers to prioritise oils such as olive oil, with butter or tallow as "other options".
On fat, the US says "More high-quality research is needed", while the paper calls saturated versus unsaturated fat "one of the most consistent findings in nutrition science". The paper presents much more evidence for its position. That doesn't make the whole pyramid settled. Its red meat case is mainly cohort and substitution analysis, plus trials on LDL, and it says associations are "generally stronger for processed meat than for unprocessed red meat".
Three corrections
USA TODAY says the US limited sugar to "10mg per meal". It's 10 grams, as the guidelines and American Family Physician both show. It says the US pyramid emphasised "trans and saturated fats", but the guidelines never mention trans fat, and we found no support for that. And the "22" scientists in its excerpt should be 23 authors: BYU's wording was one professor "alongside 22".
None of this undoes USA TODAY's central comparison, which is worth reading.