149 corrections on 15 pages. Most arrive together: one paper read again against every page that cites it, and every page fixed the same day.
Lafontant et al. 202521 corrections · 3 pages
November 20, 2026 · We got it wrong
A Short said a review found no difference in fat loss between cardio, weights and both, when cardio and both lost more fat than weights
What it said
It said body-fat changes "were no different whether people did cardio, lifted weights, or combined both", and its recording said "the fat-loss difference between all of them was not significant".
What it says now
It says that in trials of 10 weeks or more, "cardio and the combination each lost about 1 kg more fat than weights alone", and that "Body fat percentage did not clearly differ between the three."
Why
Lafontant and colleagues (2025) reviewed 36 trials of cardio, weight training and the two combined. In trials of at least 10 weeks, cardio and the combination each reduced fat mass by about 1 kg more than weight training alone; only body fat percentage showed no clear difference between the three. The Short told the review as finding no difference in fat loss at all, which the review contradicts.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights claim gave training advice no paper on it tested
What it said
It said "adding two or three resistance sessions per week won't slow your fat loss by any measurable amount", that "Even two 20-minute sessions of the other modality per week" would do, that "That advantage disappears if you shorten your rest periods during weight training", and "Shorter rests, supersets, or circuits close the gap".
What it says now
It says the review "did not test how many sessions of each you need", and gives the authors' own suggestion: "they suggest supersets and circuits" to fit the same work into less time.
Why
No paper on the page tested how many sessions a week of each kind are needed, any dose, or shorter rests. Lafontant and colleagues (2025) suggest supersets and circuits to fit the work into less time, and advise including weight training when keeping or building fat-free mass matters.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights claim described a review's untrained participants as people new to exercise or coming back after a break
What it said
It said 33 of 36 studies "used people who were new to exercise or coming back after a break", and "These results describe what happens when someone like you starts moving".
What it says now
It says those studies "tested people who were not trained, so the results fit beginners best".
Why
Lafontant and colleagues (2025) marked the participants of 33 of their 36 studies as untrained. The paper does not say any were coming back after a break, and it could not test trained people separately, as only three studies included them.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights study page told readers who want to keep muscle that doing both was the safer bet, against the authors' advice to keep weights in
What it said
Its card for readers keeping muscle said of doing both "That finding is robust", and "the safer bet in the data isn't weights-only".
What it says now
It says the authors "advise keeping weights in if lean mass matters to you", and ends "So keep weights in the plan, alone or with cardio."
Why
Lafontant and colleagues (2025) conclude that weight training alone kept more fat-free mass than cardio alone, while doing both did not clearly differ from either, and they advise including weight training when fat-free mass matters. They add that doing both may keep less fat-free mass than weights alone, a practical but not statistical difference. Doing both was not the safer bet.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights study page got two comparisons wrong: a factor of eight and a gap three times larger
What it said
It said the fitness figure for muscle was "overestimated by a factor of eight", and in women "The trend for body mass loss was nearly three times larger", which the authors "called this too small a sample to be meaningful".
What it says now
It says the figure "is 5 to 8 times too high", and that the women's result is "a gap in body weight, not fat".
Why
Thirty to 50 calories a pound against about 6 is 5 to 8 times. The women-only trend in the review by Lafontant and colleagues (2025) was about 2.9 kg of body weight, set against a 1 kg gap in fat mass; the matching body-weight gap was 1.82 kg, and the trend was not a clear difference. The authors' remark about too few studies was made of other analyses by sex.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights study page named a laboratory and a country, and described the funder, though our copy of the paper shows none of these
What it said
It named a laboratory at the "University of South Florida", gave a country, and called the funder "an educational nonprofit" with "no commercial interest".
What it says now
It says "Not shown in our copy of the paper".
Why
The copy of the review by Lafontant and colleagues (2025) that we hold folds its author information, so it shows no laboratory, university or country. It names the Florida Education Fund as its funder and says nothing more about it.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights study page said research shows weight training produces fat loss on its own, when the review only compared it with other exercise
What it said
Asked whether you can lose weight with just weight training, it answered "The research says yes — weight training does produce fat loss on its own."
What it says now
It says the review "never compared weight training with doing no exercise", and that weights alone lost "about 1 kilogram less than cardio or doing both" in trials of 10 weeks or more.
Why
Every trial in the review by Lafontant and colleagues (2025) compared weight training with cardio and with the two combined; none had a group that did not exercise. Weights alone lost less fat mass than the other two, so the review cannot say what weight training does on its own.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights study page said doing cardio and weights in the same session lost more fat, when the review compared the same day with different days
What it said
It said "doing both in the same session produced more fat loss" and "Same-session combining appears slightly favorable for fat loss".
What it says now
It says doing both "on the same day lost more fat" than weights alone, with one study of different days, and that the authors conclude "scheduling made no meaningful difference".
Why
Lafontant and colleagues (2025) compared doing cardio and weights on the same day with doing them on different days of the week, not in one session. Same-day training lost more fat mass than weights alone, but only one study tested different days, and the authors conclude that scheduling had no meaningful effect, so personal preference can decide.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights study page said six things no paper on it found, among them that in short programs only effort mattered
What it said
It said that in shorter trials "total effort during each session was the only variable that mattered", that "The percentage needle barely moved", that "most participants were younger adults", that "CT may show advantages" in longer programs, that "the broader research on spot reduction suggests it doesn't work that way", and defined trained people by "3+ years consistent training".
What it says now
It says that below ten weeks "the review found no clear difference between the types", and that "the review did not test whether age changed the results".
Why
Lafontant and colleagues (2025) found no clear difference between the types in trials under 10 weeks and said nothing about effort there. They reported no difference in body fat percentage between the types, did not split their results by age, and gave no definition of trained people in years. No paper on the page tested longer programs of doing both or spot reduction.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages said every difference vanished when the work was matched, from 12 trials the authors ask readers to treat with caution
What it said
The study page said "every single body composition difference vanished", that the trials "forced both groups to do equal amounts of total work", and called it the "work-matched nullification"; the claim said "every body composition difference disappeared entirely", "match the effort and nothing else matters" and "the gap was zero".
What it says now
They say "Twelve of the 36 studies tried to match the workload between groups", most of them by session length, that the three types "showed similar changes", and that the authors "ask readers to interpret this cautiously".
Why
Lafontant and colleagues (2025) report that in the 12 of their 36 trials that attempted to equate workloads, nine by session length, cardio, weights and both showed similar changes, with no significant differences. They ask that this smaller analysis be interpreted cautiously, because few trials qualified and they matched the work in different ways; no difference was shown to be zero. Two earlier corrections fixed part of this: 2026-11-16-cardio-vs-weights-study-page-said-differences-disappeared-entirely, whose words still stood in other parts of the study page, and 2026-11-13-cardio-vs-weights-claim-gave-a-12-trial-result-to-36-trials, which left the claim saying the 12 trials equalized total work.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages called their least certain result the most robust, and said the core finding held in every analysis
What it said
The study page said "The work-matched finding is the most robust conclusion" and "the core conclusion survived every analysis"; the claim said "The consistency of the null finding across varied matching methods strengthens it" and "the core finding held across every population, every age group, and every method".
What it says now
The study page says "The firmest result is cardio's fat-mass gap over weights alone" and "The work-matched finding is small and cautious"; the claim says the authors' caution is there "because it rests on a few trials that matched the work in different ways".
Why
Lafontant and colleagues (2025) ask readers to interpret the effort-matched result cautiously, because few trials qualified and they matched the work in different ways, and they added that analysis after registering the review. In their leave-one-out checks, no single study changed cardio's fat-mass gap over weights alone; leaving out single studies changed several other comparisons, among them doing both against weights for fat mass, and a second review found effects that varied with age and program length.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages said exercise type does not matter and that rest periods explain cardio's edge, which the review only suggests
What it said
The study page said "It was arithmetic", "Exercise type matters far less than total effort", that cardio's edge was "explained by rest periods, not biology", "The gap is largely explained by rest periods", that if you shorten them "the data suggests the results converge", and "The fat loss advantage of cardio is likely explained by rest periods"; the claim said "Not because it's biologically superior. Because of rest periods", "cardio's advantage is arithmetic" and "The question was never which exercise is better for fat loss".
What it says now
The study page says "In these trials the type of exercise mattered, if modestly"; the claim says "So the type of exercise did matter here"; both give the authors' reason as theirs: cardio's bigger drop in body weight "likely comes from burning more energy" in each session.
Why
Lafontant and colleagues (2025) found that cardio and doing both reduced fat mass more than weights alone. They suggest that cardio's greater loss of body weight is likely due to more energy used in each session, as rest periods could mean less work, and they never say that exercise type does not matter.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages told an illustration of a 45-minute weight session as what the trials measured
What it said
The claim said a 45-minute session involves "roughly 20 minutes of actual lifting and 25 minutes of standing between sets", that "Both people went to the gym for 45 minutes", and "A typical 45-minute weight session had about 20 minutes of actual movement"; the study page said "roughly 20 minutes of that session involves actual muscular effort" and its recording "Maybe twenty minutes".
What it says now
Both now say "That picture is ours", and that "the trials did not time anyone's rests".
Why
The review by Lafontant and colleagues (2025) gives no session timings, set lengths or rest lengths; it says only that rest periods could mean less work in each session. The minutes were FitChef's own illustration, and the study page's arithmetic on them was also wrong.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages gave a 1 kg gap found in trials of 10 weeks or more to all 36 trials, and told the trials as 36 research teams
What it said
The study page said "One kilogram of fat mass across the entire analysis", "about one kilogram across 36 studies", "The gap across 1,564 people?", "about one kilogram of fat mass across the full analysis" and "Thirty-six research teams", working "across four continents"; the claim said "the largest analysis of this question pooled 36 randomized trials" and "Across 36 randomized trials and 1,564 people, the fat loss difference between cardio, weights, and doing both was approximately one kilogram".
What it says now
The study page says "About one kilogram of fat mass in trials of ten weeks or more" and "This review gathered 36 trials, 31 of them pooled"; the claim says "In the trials that ran 10 weeks or more, cardio lost about one kilogram more fat than weights alone".
Why
Lafontant and colleagues (2025) gathered 36 trials and pooled 31. The gaps of about 1 kg, cardio against weights and doing both against weights, come from the trials that lasted at least 10 weeks. Two reports in the review came from one study, the paper names no continents, and it never calls itself the largest analysis. The claim's statement was corrected on 13 November (2026-11-13-cardio-vs-weights-claim-gave-a-12-trial-result-to-36-trials) but still gave the 1 kg to all 36 trials; the study page's correction of 16 November (2026-11-16-cardio-vs-weights-study-page-said-differences-disappeared-entirely) took "across 36 studies" off some of its lines but not all.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages said doing both matched cardio for fat loss and weights for muscle with no trade-off, which the review did not find
What it said
The study page said doing both "matched cardio for fat loss and matched weights for muscle preservation", that "combining both costs you nothing" and "Zero trade-off when you combine both"; the claim said "As genuine equals on both fronts", "It doesn't exist in the data" and "with zero trade-off".
What it says now
They say doing both "did not clearly differ from cardio for fat loss, or from weights for lean mass"; the claim adds "No clear difference is not the same as equal, though", and the study page "Doing both did not clearly cost either one", each with the authors' caution about lean mass.
Why
In the review by Lafontant and colleagues (2025), doing both did not differ significantly from cardio for fat mass or from weights for fat-free mass. That is not the same as equal: the authors note that doing both may keep less fat-free mass than weights alone, a practical but not statistical difference. The study page's correction of 16 November (2026-11-16-cardio-vs-weights-study-page-said-differences-disappeared-entirely) had taken the first of these phrases off some of its lines; others still carried it.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages called a review's fat-free mass muscle
What it said
The study page said weights "preserved slightly more muscle" and titled a card "Muscle preservation is your main concern"; the claim said "It protects the muscle that shapes how you actually look".
What it says now
The study page says "Lean mass counts muscle plus water and stored fuel"; the claim says "Lean mass here means everything but fat: muscle, but also water and stored fuel".
Why
Lafontant and colleagues (2025) measured fat-free mass, which also includes water and stored glycogen, as they note. The pages now say lean mass, explained in plain words, and keep the word muscle for papers that measured muscle.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages said leaving out one study erased or reversed weights' lean-mass edge, when it only stopped being clear
What it said
The study page said one study "could erase it", that "Six individual studies each independently flip the result", and in its recording "it flips"; the claim said single studies "individually reverse the finding", and both said "the advantage disappears".
What it says now
The study page says "It does not reverse; it just stops being a clear difference"; the claim says "That edge was no longer a clear difference when any one of six trials was left out".
Why
In the review by Lafontant and colleagues (2025), weights kept about 0.88 kg more fat-free mass than cardio. Leaving out any one of six studies made that result no longer statistically significant; it never reversed, and the authors still conclude that weights alone kept more fat-free mass than cardio alone.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages told a clear 1 kg gap in fat loss as too small to matter
What it said
The study page said "a gap so small you would need lab equipment to confirm it exists", "cardio barely edges weights" and "trivially small relative to the cultural debate"; the claim said "The internet's most passionate fitness argument has been about the weight of a bag of sugar" and "the fat loss difference between exercise modes is trivially small".
What it says now
The study page says "a gap that is real but modest"; the claim says "That is a real gap, if a modest one".
Why
In the review by Lafontant and colleagues (2025), cardio and doing both each lost about 1 kg more fat mass than weights alone in trials of 10 weeks or more, both clear differences, and the authors conclude they are more effective than weights alone for reducing fat mass. The paper makes no ruling that the gap is trivial.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages said diet was not controlled in most studies, when the review could not account for it
What it said
The study page said "Diet was not controlled" and "Most of the 36 studies didn't track what participants ate"; the claim said "diet wasn't controlled in most studies".
What it says now
The study page says "What participants ate was reported too unevenly across studies to control for"; the claim says "What people ate was reported too unevenly to account for".
Why
Lafontant and colleagues (2025) write that they could not control for differences in diet because studies reported it inconsistently, and that the studies which set a diet gave every group the same one. The paper does not say most studies left diet uncontrolled or untracked.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights claim and a Short told no clear difference in body fat percentage as zero or equivalent
What it said
The claim said of body fat percentage that "there was zero measurable difference between any exercise type", and that "Shorter programs showed no gap at all"; the Short said the three produced "equivalent results across 36 randomized trials".
What it says now
The claim says "the review found no clear difference between the three" and "In shorter programs, the review found no clear difference"; the Short says "Body fat percentage did not clearly differ between the three."
Why
Lafontant and colleagues (2025) found no significant difference in body fat percentage between cardio, weights and both, pooled from 31 of their 36 trials, and none between the types in trials shorter than 10 weeks. No significant difference is not zero or equivalence: it means the review could not show one.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages cut words from a quote, moved the authors' caution to another result, and set their own lines as the papers' quotes
What it said
The study page quoted the authors as saying rest could mean "less work being done during each exercise session" with words taken out, said "The researchers cautioned that results" should be read cautiously of a result the caution was not about, and set its own line "The metabolic furnace the fitness industry sold you runs on a single apple" under the Wang paper's name; the claim's source quote, "When exercise workloads were matched between conditions, no differences in any body composition outcome were observed", is not the authors' wording.
What it says now
The study page quotes "less work being done (i.e. energy expenditure) during each exercise session", and its pull quotes are labelled as FitChef's summaries; the claim's source quote for AI systems is now one of the authors' own sentences, word for word.
Why
Lafontant and colleagues (2025) wrote that rest periods could potentially result in less work being done (i.e. energy expenditure) in each session. Their call to interpret results cautiously was about the effort-matched trials, not the lean-mass result it was set beside. The pull quotes were FitChef's own summaries, and the claim's source quote put the authors' finding in other words.
Lafontant K, Rukstela A, Hanson A, Chan J, Alsayed Y, Ayers-Creech WA, Bale C, Ohigashi Y, Solis J, Shelton G, Alur I, Resler C, Heath A, Ericksen S, Forbes SC, Campbell BI. Comparison of concurrent, resistance, or aerobic training on body fat loss: a systematic review and meta-analysis. J Int Soc Sports Nutr. 2025;22(1):2507949.Read the paper
Found by
Our full re-read of the page against all its papers.
Guo et al. 20237 corrections · 2 pages
November 20, 2026 · We got it wrong
A Short gave a review's body-weight gap as the fat-loss gap between walking and running, which no trial in it compared
What it said
Its title said "A Third of a Kilogram", and it said "the fat-loss difference between running-pace and walking-pace cardio" was less than a third of a kilogram.
What it says now
Its title is "Walking vs Running for Fat Loss: Hard vs Steady Cardio", and it says "Body weight fell by about the same amount with both kinds of cardio" and "None of them put walking against running."
Why
Guo and colleagues (2023) pooled 29 trials of hard interval cardio against steady, moderate cardio, mostly on exercise bikes; none compared walking with running. The third of a kilogram was the gap in body weight, not in fat, and it was not a clear difference.
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A Short said hard and steady cardio gave virtually identical fat loss, leaving out that intervals lowered body fat percentage and waist size more
What it said
It said "Walking and running produce virtually identical fat loss in controlled trials", that the body-fat percentage gap "told the same story", and that "The choice between walking and running was never about burning more fat."
What it says now
It says "Hard intervals did come out a little ahead on two measures" and "Body fat percentage fell by about half a point more, and waist size by about 1 cm more."
Why
In the review by Guo and colleagues (2023), interval training lowered body fat percentage by about half a point more than steady cardio, and waist size by about 1 cm more; weight and fat mass did not clearly differ. The authors call these gains limited. The Short told the trials as showing no difference in fat at all.
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A Short named the wrong journal for a cardio review and set its own summary under the authors' names
What it said
It named the review's journal as Sports Medicine, set its takeaway under "— Guo et al. 2023 · Sports Medicine · 29 RCTs", and cited the two reviews as "HIIT vs MICT for Body Composition (29 RCTs)" and "Exercise Modality and Body Fat (36 RCTs)".
What it says now
It names the journal as "International Journal of Environmental Research and Public Health", marks the takeaway as "FitChef summary of Guo et al. 2023", and cites each review by its own title.
Why
The review by Guo and colleagues (2023) was published in the International Journal of Environmental Research and Public Health, not Sports Medicine. The takeaway's sentences are FitChef's summary, not the authors' words, and the two citation names were labels FitChef composed.
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A Short gave a weight result as coming from 29 trials and copied a review's misprinted range
What it said
It said "Across 29 controlled trials" for a body-weight result from 28 trials, gave the range "−0.86 to −0.26", and called a 5 percent mark "their own threshold".
What it says now
Its researcher notes give the weight gap from "28 studies", leave the misprinted range out of the figures, and say the 5 percent mark came from "the threshold of an earlier paper".
Why
Guo and colleagues' body-weight result pooled 28 of their 29 trials, and they printed its 95% range as −0.86 to −0.26, which cannot hold with a mean of −0.32 and p = 0.2514. The 5 percent mark for a meaningful drop in body fat came from an earlier paper they cite (Stevens and colleagues, 2006).
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A Short said that because walking and running sit inside the trials' intensity range, their gap must be even smaller
What it said
It said "walking and running fall within that intensity range" and "the narrower walking-to-running gap within it produces even less", and its researcher notes called the gap "an upper bound".
What it says now
It says "for walking against running, the trials here are only a rough guide".
Why
Guo and colleagues (2023) compared interval training at 80 to 100 percent of maximum heart rate with steady training at 40 to 80 percent. No trial compared walking with running, and no paper here tested how the gap changes between them; the smaller gap was FitChef's inference, told as a finding.
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A Short said three things no paper on it showed: that the treadmill's calorie number is accurate, that this training builds muscle, and that the body enforces a calorie ceiling
What it said
It said "The calorie-per-minute number on your treadmill is accurate", that "training builds muscle", and "Your body enforces a ceiling on exercise calories".
What it says now
It no longer says the first two, and it links the calorie question as "a question of its own".
Why
No paper on the page tested treadmill displays. In the review by Guo and colleagues (2023) neither kind of cardio changed lean mass, so it showed no muscle built; and Pontzer and colleagues (2016) saw daily energy use level off across people at one point in time, not a ceiling each body enforces.
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights claim said cardio does not keep lean mass, when a review found lean mass held about level with either kind of cardio
What it said
It said aerobic exercise "does not build or preserve lean mass", that "cardio alone doesn't build or preserve muscle", and that "Only adding resistance changed that".
What it says now
It says that in that review "fat-free mass was kept, not built", and that "neither moderate nor hard cardio changed lean mass clearly".
Why
In the review by Guo and colleagues (2023), fat-free mass did not change clearly with either interval or steady cardio: it was neither lost nor gained. That shows the cardio did not build lean mass, not that it failed to keep it. Guo and colleagues tested no weight training, so their review could not show that adding weights changed anything.
Guo Z, Li M, Cai J, Gong W, Liu Y, Liu Z. Effect of High-Intensity Interval Training vs. Moderate-Intensity Continuous Training on Fat Loss and Cardiorespiratory Fitness in the Young and Middle-Aged a Systematic Review and Meta-Analysis. Int J Environ Res Public Health. 2023;20(6):4741.Read the paper
Found by
Our full re-read of the page against all its papers.
Pontzer et al. 20162 corrections · 1 page
November 20, 2026 · We got it wrong
A Short gave one study's idea of how the body offsets exercise as the reason walking and running burn similar fat
What it said
It said the body "compensates for it, so it never translates to meaningful extra fat lost", "pulling energy from inflammation, immune function, and other background processes to absorb the extra workout burn".
What it says now
It says "The authors think the body may make up for it by spending less energy elsewhere, and say the details are not yet clear."
Why
Pontzer and colleagues (2016) compared 332 adults at one point in time and found that daily energy use leveled off in the most active. Why it leveled off is their hypothesis, which they could not test, and they did not study walking, running or fat loss. No paper links that plateau to the trials of interval against steady cardio.
Pontzer H, Durazo-Arvizu R, Dugas L, Plange-Rhule J, Bovet P, Forrester TE, Lambert EV, Cooper RS, Schoeller DA, Luke A. Constrained Total Energy Expenditure and Metabolic Adaptation to Physical Activity in Adult Humans. Curr Biol. 2016;26(3):410-417.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A Short gave activity's 7 to 9 percent as a share of all differences in daily calorie burn, and the rest to body size
What it said
It said "Physical activity explains only 7 to 9 percent of the difference" and "Body size drives nearly everything else", and its chart gave about 93 percent to body size.
What it says now
It says "Once body measures and place were taken into account, activity explained only 7 to 9 percent of what was left", and its chart is labelled "Not explained by activity".
Why
Pontzer and colleagues (2016) found that body measures explained just over half of the differences in daily energy use between people. The 7 to 9 percent was the share activity explained of what was left once body measures and place were set aside, not a share of the whole, and the rest was not body size.
Pontzer H, Durazo-Arvizu R, Dugas L, Plange-Rhule J, Bovet P, Forrester TE, Lambert EV, Cooper RS, Schoeller DA, Luke A. Constrained Total Energy Expenditure and Metabolic Adaptation to Physical Activity in Adult Humans. Curr Biol. 2016;26(3):410-417.Read the paper
Found by
Our full re-read of the page against all its papers.
Khalafi et al. 20253 corrections · 2 pages
November 20, 2026 · We got it wrong
A cardio versus weights claim said the benefit of doing both grows with age, when a review found it tended to be smaller after 65
What it said
It said "The older you are, the more the evidence supports doing both", that combining "may matter even more as you age", that "That wasn't true for younger groups", and answered whether the evidence applies past 40 with "Yes — directly."
What it says now
It says the gains from doing both "tended to be bigger in middle-aged adults than in those 65 and older", and that the 36-trial review "did not check whether age changed the results".
Why
Khalafi and colleagues (2025) reviewed 53 studies in adults aged 50 and over. They found the effects of doing both tended to be greater in middle-aged adults than in adults 65 and older, and wrote that middle-aged adults may be more likely to benefit. No paper on the page compared younger adults or people over 40, and the 36-trial review by Lafontant and colleagues did not test age.
Khalafi M, Kheradmand S, Habibi Maleki A, Symonds ME, Rosenkranz SK, Batrakoulis A. The Effects of Concurrent Training Versus Aerobic or Resistance Training Alone on Body Composition in Middle-Aged and Older Adults: A Systematic Review and Meta-Analysis. Healthcare (Basel). 2025;13(7):776.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cardio versus weights claim gave four numbers and checks its reviews do not give
What it said
It said "That difference was clear across 53 studies", gave a fat-mass result as "p = 0.919", and described the main reviews with "I2 = 0% for most between-group comparisons in both Lafontant and Guo" and "Low per funnel plots and Egger's test in both primary meta-analyses".
What it says now
It gives the lean-mass difference for adults 50 and over "from 27 trials", and its researcher notes give each review's checks as the review reports them.
Why
In the review by Khalafi and colleagues (2025), the lean-mass comparison of doing both against cardio pooled 27 trials, not 53, and the fat-mass result had p = 0.91; 0.919 came from a different meta-analysis. The text of the review by Lafontant and colleagues reports no heterogeneity figures and inspected funnel plots only, with results in an appendix we do not hold, while Khalafi and colleagues found signs of publication bias for several outcomes.
Khalafi M, Kheradmand S, Habibi Maleki A, Symonds ME, Rosenkranz SK, Batrakoulis A. The Effects of Concurrent Training Versus Aerobic or Resistance Training Alone on Body Composition in Middle-Aged and Older Adults: A Systematic Review and Meta-Analysis. Healthcare (Basel). 2025;13(7):776.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages told a second review as independent confirmation and added up their trials, though the reviews share at least ten
What it said
The study page said the second review was "arriving at the same answer independently", that it "confirmed the same pattern", and counted "= 89 studies"; the claim said "Three independent meta-analyses converge", "Zero divergent findings", counted "118 underlying trials and 5,244 participants", and rated its certainty high.
What it says now
The study page says "It shares at least ten trials with this one"; the claim says "At least ten of its trials also appear in the first review" and rates the evidence "Moderate Certainty".
Why
The review by Khalafi and colleagues (2025), of 53 studies in adults aged 50 and over, shares at least ten trials with the review by Lafontant and colleagues, so the two are not independent and their counts cannot be added. Khalafi and colleagues found doing both as effective as cardio for body fat and as weights for building muscle, with effects that varied with age and program length, but no clear fat-mass difference between doing both and weights alone, where Lafontant and colleagues found one. The claim's third review, by Guo and colleagues, compared two kinds of cardio, not cardio and weights.
Khalafi M, Kheradmand S, Habibi Maleki A, Symonds ME, Rosenkranz SK, Batrakoulis A. The Effects of Concurrent Training Versus Aerobic or Resistance Training Alone on Body Composition in Middle-Aged and Older Adults: A Systematic Review and Meta-Analysis. Healthcare (Basel). 2025;13(7):776.Read the paper
Found by
Our full re-read of the page against all its papers.
A cardio versus weights claim said again that none of 15 studies found a cost to muscle growth, which its source does not report
November 20, 2026 · We got it wrong
What it said
It said of the 15 studies that measured muscle growth "and none of those studies reported a significant cost".
What it says now
It gives the pooled result alone: "the pooled difference was too small to tell apart from zero".
Why
Schumann and colleagues (2022) pooled 15 studies of muscle growth and found no interference from adding cardio. Single studies ran from slightly lower to slightly higher growth, and the paper does not report whether any one of them found a significant cost. The correction 2026-10-13-concurrent-training-meta-analysis-told-as-zero-and-unanimous took the words that every study agreed off this page; this clause said it again in other words.
Schumann M, Feuerbacher JF, Sünkeler M, Freitag N, Rønnestad BR, Doma K, Lundberg TR. Compatibility of Concurrent Aerobic and Strength Training for Skeletal Muscle Size and Function: An Updated Systematic Review and Meta-Analysis. Sports Med. 2022;52(3):601-612 (online 10 November 2021).Read the paper
Found by
Our full re-read of the page against all its papers.
Wang et al. 20102 corrections · 2 pages
November 20, 2026 · We got it wrong
A cardio versus weights study page gave two papers wrong titles, one a wrong journal and a wrong author
What it said
It cited the Wang paper as "Specific metabolic rates of major organs and tissues across adulthood" in the "Journal of Applied Physiology", listed "Schautz B" among its authors, and gave the second review the title "Effects of concurrent training on body composition in older adults".
What it says now
It cites "Wang et al., American Journal of Human Biology" as "the authors' accepted manuscript, PMC3139779", and each paper by its own title.
Why
The Wang paper we hold is the authors' accepted manuscript of a study in the American Journal of Human Biology (PMC3139779); its fifth author is Martin Heller. The second review, by Khalafi and colleagues (2025), is titled The Effects of Concurrent Training Versus Aerobic or Resistance Training Alone on Body Composition in Middle-Aged and Older Adults. A note for AI systems also set the two reviews' authors beside one review's link; it now names Lafontant and colleagues alone.
Wang Z, Ying Z, Bosy-Westphal A, Zhang J, Heller M, Later W, Heymsfield SB, Müller MJ. Evaluation of Specific Metabolic Rates of Major Organs and Tissues: Comparison Between Men and Women. Am J Hum Biol. 2010 Dec 22;23(3):333-338 (the authors' accepted manuscript, PMC3139779).Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two cardio versus weights pages said a study measured muscle's resting burn and tested the fitness figure, when it checked an earlier estimate
What it said
The study page said "Wang and colleagues tested that claim directly", that "MRI measurement shows 6 calories", and gave the study's ages as "Adult (specific range in paper)"; the claim said "The measured number, from researchers who scanned actual organs" and named the source "Organ-level metabolic scanning".
What it says now
The study page says "Wang and colleagues did not test the fitness figure" and "The 1992 values held for both men and women"; the claim names the study "of 106 healthy adults aged 20 to 49", which "checked a 1992 estimate for resting muscle".
Why
Wang and colleagues checked, in 106 healthy adults aged 20 to 49 who were not obese, whether a 1992 estimate of each organ and tissue's resting energy use held for men and women. For muscle it is 13 calories per kilogram a day, about 6 per pound, and it held. MRI measured the size of organs and tissues; resting energy use came from breathing measurements. The paper never mentions the 30 to 50 calories a pound often quoted in fitness. We hold the authors' accepted manuscript (PMC3139779), and the pages now name it.
Wang Z, Ying Z, Bosy-Westphal A, Zhang J, Heller M, Later W, Heymsfield SB, Müller MJ. Evaluation of Specific Metabolic Rates of Major Organs and Tissues: Comparison Between Men and Women. Am J Hum Biol. 2010 Dec 22;23(3):333-338 (the authors' accepted manuscript, PMC3139779).Read the paper
Found by
Our full re-read of the page against all its papers.
Two cardio versus weights pages described linked pages with findings their papers do not make
November 20, 2026 · We got it wrong
What it said
The study page said "131 studies across three separate questions converged on the same answer", that 43 studies "reached the same conclusion" for muscle growth, and that a model "tested that assumption directly"; the claim said exercise mode becomes "progressively less relevant", that "load doesn't determine muscle growth", that "sex doesn't determine hypertrophy rate", and linked "nine variables tested across seven research teams and 9,350 participants".
What it says now
The study page links "our cardio guide" and says the 43-study analysis "looked at muscle growth in 15 of them"; the claim says "Neither study compared exercise types", that "absolute gains in muscle size slightly favoured men", and links "our training guide".
Why
Each line now says what the linked page covers. Schumann and colleagues (2022) measured muscle growth in 15 of their 43 studies. Pontzer and colleagues (2016) compared people who moved less and more at one point in time, and neither they nor Careau and colleagues (2021) compared exercise types. Schoenfeld and colleagues (2017) found similar growth with light and heavy loads when every set was taken to failure, and Refalo and colleagues (2025) found that absolute gains slightly favoured men while relative gains were similar.
Our full re-read of the page against all its papers.
Li et al. 202436 corrections · 2 pages
November 20, 2026 · We got it wrong
A cooled-rice claim said the weight came back after the supplement ended, without naming the group; the trial saw that in the group that took it first, the only group with a break after it
What it said
Its evidence card for the weight-loss trial said "Weight regained during 4-week washout period after RS2 supplementation ended", and its numbers for researchers said "Full weight regain within 4 weeks".
What it says now
They say "In the group that took RS2 first, weight was regained during the 4-week washout period" and "In the group that took RS2 first (RS-Washout-CS, n = 19): full weight regain by the end of the 4-week washout".
Why
The trial (Li and colleagues, Nature Metabolism, 2024) was a crossover: one group took the resistant starch first and the other the control starch first, with a four-week break between the two eight-week periods. Only the group that took the resistant starch first had that break after it, and in that group weight and the other obesity measures returned to their starting levels. The corrections of 20 November named that group on the resistant starch study page, in its story (2026-11-20-resistant-starch-study-said-the-weight-came-back-the-moment-the-supplement-stopped) and in its findings (2026-11-20-resistant-starch-study-findings-said-the-weight-came-back-when-participants-stopped-the-supplement). The same day this claim's verdict and answers were rewritten to name it too, but these two lines were left.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said you cannot easily check whether you carry a gut bacterium, and that acting on it is hardest; no paper it rests on tested either
What it said
Its story said "You can’t easily check whether you have it." and called the finding that the starch works at least partly through the bacterium "the hardest for any one person to act on".
What it says now
The story no longer says either. The page's answer on testing for the bacterium says "None of the studies on this page tested home gut-bacteria kits".
Why
The trial (Li and colleagues, 2024) found the bacterium by sequencing the DNA of the gut bacteria in stool samples. No paper this page rests on tested whether, or how easily, a person can check for it, or what anyone could do about carrying it. The correction of 20 November (2026-11-20-resistant-starch-study-judged-home-gut-bacteria-tests) took the claims about home test kits off the page's answer on testing, but left these two lines in its story.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study called the diet the condition the viral version never mentions; the trial gave everyone the same diet and never tested it as a condition
What it said
Its story introduced three earlier trials with "Then there’s the condition the viral version never mentions."
What it says now
It says "This trial was not the first to give people resistant starch." and, after the earlier trials, "The authors think the diet may be one reason their trial found weight loss where those did not."
Why
In the trial (Li and colleagues, 2024), everyone ate the same diet the researchers provided while taking either starch, so it never tested the diet as a condition of the weight loss. The authors suggest the diet may be one reason earlier trials found no weight loss. The correction of 20 November (2026-11-20-cooled-rice-pages-said-the-supplement-only-worked-with-a-controlled-diet) gave the diet as the authors' suggestion in the paragraph after this line, but left the line that called it the condition.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study's findings said the starch had no effect in mice with no gut bacteria and that one bacterium restored its benefit; the mice given the bacterium weighed less than those without it
What it said
Its findings said "the resistant starch had no effect — but adding one species of bacteria restored the benefit", and its notes for AI systems said "Weight loss occurred only when one specific bacterial species (B. adolescentis) was present."
What it says now
Its findings say "resistant starch made no clear difference to their weight" and "Mice on the same diet that were also given one kind of gut bacteria weighed less than those without it." Its notes for AI systems say the same.
Why
In the paper (Li and colleagues, 2024), mice with no gut bacteria ate a diet with resistant starch or with ordinary starch, and some were also given B. adolescentis. The starch alone made no significant change in weight, blood sugar control or insulin sensitivity, and the paper reports more gut contents after it. The mice also given the bacterium weighed less than mice on the same starch without it. The correction of 20 November (2026-11-20-resistant-starch-study-told-the-germ-free-mouse-experiment-wrong) fixed the story, the pull quote and the takeaways, but left these two lines.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study's findings said the improvements reversed when participants stopped the supplement; the trial saw that in the group that took it first, the only group with a break after it
What it said
Its findings said "When participants stopped taking the supplement, all weight-related improvements reversed within four weeks." Their statement for researchers and its notes for AI systems told the return to starting levels without naming the group.
What it says now
They say "In the group that took the supplement first, the weight and fat came back within four weeks of stopping it." Their statement for researchers and its notes for AI systems name that group too.
Why
The trial (Li and colleagues, 2024) was a crossover: one group took the resistant starch first and the other the control starch first, with a four-week break between the two periods. Only the group that took the resistant starch first had that break after it, and in that group weight and the other obesity measures returned to their starting levels. The correction of 20 November (2026-11-20-resistant-starch-study-said-the-weight-came-back-the-moment-the-supplement-stopped) named that group in the story, but left these lines.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study told AI systems how carriage of a gut bacterium varies and how well it is known across populations; no paper it rests on says either
What it said
Its notes for AI systems said "B. adolescentis prevalence varies by population, geography, diet, and baseline microbiome." and "B. adolescentis prevalence in the general population is not well-characterized across diverse populations".
What it says now
Its notes for AI systems no longer say either. They keep "Mouse mechanisms may not directly translate to human physiology." and give the authors' own limit: "The authors note that the broader generalizability of their results requires additional validation in larger and more diverse cohorts."
Why
No paper this page rests on gives how common B. adolescentis is in any group of people, what that depends on, or how well it is known. The trial (Li and colleagues, 2024) counted it only in its own participants: 9 of the 27 whose gut bacteria were sequenced; its authors write that their results need checking in larger and more diverse groups of people. The correction of 20 November (2026-11-20-cooled-rice-pages-gave-how-many-people-carry-a-gut-bacterium-from-no-paper) took the carrier rates off the page, but left these lines.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said a supplement's weight loss depended on one gut bacterium; the trial found it worked at least partly through it, and carriers lost more fat
What it said
The cooled-rice claim said "Without this bacterium, the supplement did nothing — even at the full dose." and "only participants who carried this bacterium lost weight". The resistant starch study said "the effect required specific gut bacteria".
What it says now
The claim says "The trial's authors traced the weight loss at least partly to a gut bacterium called B. adolescentis." and "Those who carried it lost more body fat on the supplement than those who did not." The study says "The authors tie the effect at least partly to one gut bacterium".
Why
The trial's authors (Li and colleagues, Nature Metabolism, 2024) write that resistant starch can help weight loss at least partly through one gut bacterium, B. adolescentis. Of the 27 participants whose gut bacteria were tested, the 9 who carried it at the start lost more fat than the others; the paper never says the others lost nothing. The result where the starch alone made no difference was in mice with no gut bacteria at all. Our notes turned the authors' "at least partly" into a requirement.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study gave how many people carry a gut bacterium, by body weight, from no paper they cite
What it said
The claim said "Population data suggests 60 to 80 percent of healthy-weight adults carry it." and showed rates by weight in a chart credited to the trial. The study said "Research suggests the species is present in 60 to 80 percent of healthy-weight adults."
What it says now
They give the trial's own count. The claim says "Of the 27 participants whose gut bacteria were tested, only about a third had it at the start." and its chart shows that count. The study says "Those 9 lost more body fat on the starch than the 18 who did not carry it."
Why
No paper these pages rest on gives a carrier rate for any group of people. The weight-loss trial (Li and colleagues, 2024) counted the bacterium only in its own participants: 9 of the 27 whose gut bacteria were sequenced carried it at the start. The rates came from our notes, not from a paper.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study called the weight-loss trial's supplement purified; it was a high-amylose maize starch
What it said
The pages said the trial used "a purified resistant starch supplement at 40 grams per day" (the claim) and "a sachet of purified corn starch powder" (the study).
What it says now
They say "a high-amylose maize starch supplement that gave 40 grams of resistant starch a day" (the claim) and "a sachet of high-amylose corn starch powder" (the study).
Why
The trial (Li and colleagues, 2024) gave 91.2 grams a day of a high-amylose maize starch, Hi-Maize 260, which held 40 grams of resistant starch. The paper never calls it purified, and the 40 grams were the resistant starch, not the powder.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said the supplement worked only with a controlled diet; the trial gave everyone that diet and never tested it without
What it said
The claim said "the weight loss only appeared when participants followed a controlled eating pattern" and that earlier trials "using the same supplement without dietary control showed zero weight loss". The study said the trial worked "because the researchers controlled the entire background diet".
What it says now
The claim says "everyone also ate a diet the researchers provided, so it never tested the supplement without it". The study says "The authors think the diet may be one reason their trial found weight loss where those did not."
Why
In the trial (Li and colleagues, 2024), everyone ate the same diet the researchers provided while taking either starch, so the diet was never tested as a condition. The authors offer it as one possible reason earlier trials found no weight loss, and write that the fat in the diet matters. The pages told that suggestion as a finding.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study credited the weight-loss trial to "Yuan et al." and gave its papers titles they do not have
What it said
The pages credited the trial to "Yuan et al.", and the claim gave it the title "Resistant starch reshapes the gut microbiota to produce weight loss in humans". The claim titled the rice trial "Effect of cooling and reheating on resistant starch content and glycemic response in rice".
What it says now
They credit "Li et al." and give each paper its own title, among them "Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota".
Why
The trial's first author is Huating Li; Rui Yuan is the ninth of its authors (Nature Metabolism, 2024). The titles the claim showed were not the papers' own. The evidence cards now give each paper's real title and DOI, and the dose charts credit their 4.4 grams to the rice trial (Strozyk and colleagues, 2022).
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim showed FitChef's notes as the weight-loss paper's own words
What it said
The quote from the weight-loss paper that the page keeps for search engines and AI systems, which the page does not show, held our own notes, ending "The bacterium is the necessary gatekeeper."
What it says now
That quote, which the page does not show, now gives the paper's own words: "RS played the role in facilitating weight loss, at least partially, through B. adolescentis".
Why
The quote held a summary from our notes, not sentences from the paper (Li and colleagues, Nature Metabolism, 2024). It now holds two of the paper's own sentences, copied from the paper as pasted.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim said resistant starch is fiber your stomach cannot break down; the papers say the small intestine
What it said
The page said resistant starch is "a type of fiber your stomach cannot fully break down".
What it says now
It says "a type of fiber your small intestine cannot fully break down".
Why
The papers describe resistant starch as starch that is not digested in the small intestine and moves on to the colon, where gut bacteria ferment it (Li and colleagues, 2024; Strozyk and colleagues, 2022).
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study described how the weight-loss trial measured its results in ways the paper does not
What it said
The claim listed the trial's methods as "(DXA, clamp, metagenomics, FMT, germ-free mice)". The study called its insulin test the "gold-standard hyperinsulinemic-euglycemic clamp" and its insulin result one "measured by the most precise method available".
What it says now
The claim's evidence notes, which the page does not print, list "(MRI, clamp, metagenomics, FMT, germ-free mice)". The study says "Significant improvement in insulin sensitivity, measured by hyperinsulinemic-euglycemic clamp".
Why
The paper (Li and colleagues, 2024) measured belly fat by MRI; it names no DXA scan, and it does not call the clamp test the gold standard or the most precise method.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study called the trial's participants Chinese adults fed hospital meals; the paper says the trial ran in Shanghai with meals the researchers provided
What it said
The study said "Thirty-seven Chinese adults in Shanghai" got "hospital-provided meals". The claim listed "Chinese adults only" among the trial's limits.
What it says now
The study says "Thirty-seven adults in Shanghai" who "ate meals provided by the researchers". The claim says the trial was "conducted in Shanghai".
Why
The paper (Li and colleagues, 2024) says the study was conducted in Shanghai and that the researchers provided three meals a day. It does not give the participants' ethnicity or say the meals came from a hospital kitchen.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said no study has tested cooled rice for weight loss; they can say only that none of their studies did
What it said
The claim said "No study has tested whether the kitchen version produces the same weight effect." The study said "No controlled trial has tested RS3 from cooled rice for weight loss."
What it says now
They say "None of these three studies tested whether the kitchen version produces the same weight effect." and "None of the studies on this page tested RS3 from cooled rice for weight loss."
Why
The papers behind these pages tested a starch supplement (Li and colleagues, 2024), blood sugar after one meal of cooled rice (Strozyk and colleagues, 2022) and blood sugar in a review of trials (Pugh and colleagues, 2023). They show that none of them tested cooled rice for weight loss, not that no study anywhere has.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study called the weight-loss trial one of the most rigorous of the decade, in one of the most respected journals; no paper says so
What it said
The claim called the trial "one of the most rigorous nutrition experiments published in the last decade", and the study said its result came "from one of the most respected journals in metabolic science".
What it says now
The claim says what made it careful: "Neither the people taking part nor the researchers knew who got which starch, and everyone ate the same diet." The study says "That is a legitimate result."
Why
Neither paper ranks the trial or the journal. The trial (Li and colleagues, 2024) was placebo-controlled and double-blinded, and everyone ate the same diet; the claim now says that instead.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said trial participants lost 2.81 kilograms, beside a placebo group; 2.81 kilograms was the gap between their two starches
What it said
The claim said "participants lost an average of 2.81 kilograms". The study said "They lost an average of 2.81 kilograms compared to a placebo group who gained half a kilo".
What it says now
The study says "On the resistant starch, their weight fell by about 2.3 kilograms; on the control starch, it rose by about half a kilo." The claim says "participants lost about 2.8 kilograms more than during eight weeks on a control starch".
Why
The trial (Li and colleagues, 2024) was a crossover: each person took the resistant starch for eight weeks and a look-alike control starch for eight weeks. On the resistant starch their weight went from 83.24 to 80.91 kilograms on average, on the control starch from 82.74 to 83.23. The 2.81 kilograms is the difference between the two, not the weight lost, and there was no separate placebo group.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study told the trial's P value as the chance its result was random
What it said
It said the gap was "so consistent the researchers calculated less than a one-in-a-thousand chance it was random".
What it says now
It says the gap "was a clear difference, very unlikely to be down to chance alone".
Why
The paper (Li and colleagues, 2024) reports P < 0.001 for the difference. That means a gap this large would be very unlikely if the starch made no difference; it is not the chance that the result was random.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the starch did nothing in mice with no gut bacteria until a bacterium was added to the same mice; separate mice were given it, and the starch alone made no clear difference
What it said
It said "Nothing happened." and "The starch passed through like any other carbohydrate.", then that the researchers "gave the same mice the same starch again".
What it says now
It says "Next to similar mice fed ordinary starch, the resistant starch made no clear difference to their weight, their blood sugar or how well their insulin worked." and "Other mice on the same resistant starch diet were also given a single bacterial species".
Why
In the paper (Li and colleagues, 2024), separate groups of mice with no gut bacteria ate a diet with resistant starch or with ordinary starch, and some were also given B. adolescentis. The starch alone made no significant change in weight, blood sugar control or insulin sensitivity; the mice also given the bacterium weighed less. The paper also reports more gut contents after the starch, so it did not simply pass through.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said a third of the trial's participants lacked the bacterium; a third had it
What it said
A quote box said "A third of the study's own participants didn't have that species."
What it says now
It says "Only a third of the 27 participants tested had that species."
Why
Of the 27 participants whose gut bacteria were sequenced (Li and colleagues, 2024), 9 carried B. adolescentis at the start and 18 did not: a third had it, two thirds did not.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said 40 grams a day had been tested for weight loss exactly once; earlier trials at about that dose found no weight loss
What it said
It said the claim "has been tested exactly once, under ideal conditions".
What it says now
It says "Earlier trials that gave about the same 40 grams a day found no weight loss."
Why
The weight-loss paper (Li and colleagues, 2024) cites an earlier trial that gave about 40 grams of resistant starch a day without weight loss, and the 2023 review (Pugh and colleagues) places trials at 40 grams a day among those with no significant weight change. The trial is the one on this page that found weight loss at that dose.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the supplement repaired the gut lining; that was found in mice given gut bacteria from people who took it
What it said
It said the starch "repaired the gut lining itself" and "The supplement repaired the gut lining".
What it says now
It says "Mice were given gut bacteria from people who had taken the supplement. Their gut wall sealed better than in mice given bacteria from people on the control starch."
Why
In the paper (Li and colleagues, 2024), the gut-wall results come from mice given gut bacteria from people after the supplement: their gut let less through, the proteins that seal it rose, and less bacterial toxin reached the blood. The paper did not measure the gut lining in the trial's participants.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said all its funding came from Chinese government programs and its authors had no commercial incentive; the paper lists more funders, and a company provided both starches
What it said
It said "All funding came from Chinese government research programs" and "The researchers had no commercial incentive to find a positive result for resistant starch products."
What it says now
Its stored integrity notes, which the page does not print, say "The paper lists no company among its funders. Both starches the participants took, the resistant starch and the control starch, were provided by Ingredion." The page now gives the paper's declaration: "The authors declare no competing interests."
Why
The paper (Li and colleagues, 2024) lists funders in mainland China and Hong Kong, the Marie Sklodowska-Curie Actions and Germany's DFG. It says both starches were provided by Ingredion, and that the authors declare no competing interests. It says nothing about the authors' incentives.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said home gut-bacteria tests detect species inconsistently; no paper it cites tested them
What it said
It said "species-level detection is inconsistent across commercial kits" and that the trial's method was one "most consumer tests do not match".
What it says now
It says "None of the studies on this page tested home gut-bacteria kits".
Why
The trial (Li and colleagues, 2024) found the bacterium by sequencing the DNA in stool samples. None of the papers this page rests on tested a home kit.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said its result applies to other East Asian populations and set people with gut disorders outside it; the paper says neither
What it said
It said the result applies to "Adults with overweight/obesity from other East Asian populations with similar dietary patterns and gut microbiome profiles" and listed "those with established gastrointestinal disorders affecting microbiome".
What it says now
It says "The authors note that applicability to other populations remains unconfirmed."
Why
The paper (Li and colleagues, 2024) says its applicability to other populations remains unconfirmed and names no gut disorder among its exclusions. The high-fat diets it mentions were fed to rodents.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study credited an earlier trial's fat-to-fibre figure to the wrong trial
What it said
It said "Authors cite Robertson 2012 (fat:fibre ~53:27 g/d), Johnston 2010, and Maki 2012 as trials without weight loss."
The paper (Li and colleagues, 2024) gives the 53:27 figure for Upadhyaya 2016 and 77:40 for Robertson 2012, and says Johnston 2010 and Maki 2012 set no background diet.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the weight came back the moment the supplement stopped; it came back over the four-week break, in the group that took it first
What it said
It said the effects "reversed the moment supplementation stopped" and "Every kilo came back."
What it says now
It says "in the group that took the starch first, the weight came back by the end of the four-week break".
Why
The paper (Li and colleagues, 2024) reports that in the group that took the resistant starch first, weight and the other obesity measures returned to their starting levels by the end of the four-week break.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the trial's participants never ate rice; the paper does not list their meals
What it said
Its summary for AI systems said "The study participants never ate rice."
What it says now
It says "The trial tested a starch supplement, not cooled rice."
Why
The paper (Li and colleagues, 2024) says the researchers provided a balanced diet, three meals a day, but does not list its foods. What it tested was a starch supplement, not cooled rice.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the gut bacteria were sequenced in 43 samples; there were 86
What it said
Its research notes said "n=43 samples: 27 RS + 16 CS".
What it says now
They say "86 samples: 27 participants before and after RS, 16 before and after CS".
Why
The paper (Li and colleagues, 2024) sequenced 86 stool samples, from 27 participants before and after the resistant starch and 16 before and after the control starch.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study graded the trial's results as clinically meaningful and said the supplement reduced inflammation; two of five inflammation markers fell
What it said
Its research panel called the waist result a "Clinically meaningful waist circumference reduction indicating visceral fat loss" and two results "Reduced systemic inflammation".
What it says now
It says "Two of the five inflammation markers measured in the blood went down; the other three did not change."
Why
The paper (Li and colleagues, 2024) reports that two blood markers of inflammation, TNF-alpha and IL-1-beta, were lower after the resistant starch and three others did not change. It does not grade its results as clinically meaningful.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the supplement increased gut bacteria linked to short-chain fatty acids; the paper tied no species to them
What it said
It said the supplement increased "species linked to starch digestion and short-chain fatty acid production".
What it says now
It says it increased "species that break down resistant starch".
Why
In the paper (Li and colleagues, 2024), the species that rose include two the authors call primary resistant starch degraders. Two short-chain fatty acids in stool fell and the others did not change, and no species was tied to them.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said gut bacteria from supplement responders transferred the weight-loss effect to mice
What it said
It said bacteria "from supplement responders into mice transferred the weight-loss effect to the new host".
What it says now
It says "Mice got gut bacteria from people who had taken the supplement."
Why
The paper (Li and colleagues, 2024) chose donors whose weight change was close to their group's average, not the best responders. The mice given their gut bacteria weighed less and had less fat than mice given bacteria from people on the control starch.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study gave the trial's entry rule as a BMI of 24 or more; people with a large waist could join too
What it said
Its participant details said "BMI >= 24 (mean 28.58 +/- 3.79)".
What it says now
They say "BMI >= 24 or waist circumference >= 85 cm (men) or >= 80 cm (women)".
Why
The paper (Li and colleagues, 2024) took adults aged 18 to 55 with a BMI of 24 or more, or a waist of at least 85 cm in men and 80 cm in women.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the trial's four-week break left only a thin margin; no paper says so
What it said
It said the washout data "suggests this assumption holds, but the margin is thin".
What it says now
It says "The washout data, with no order effect, suggests this assumption holds."
Why
The paper (Li and colleagues, 2024) found no effect of the order in which people took the two starches. It says nothing about a thin margin.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said mice given B. adolescentis were protected from weight gain on a high-fat diet; they gained less on a western diet
What it said
It said the bacterium "protected them from gaining weight on a high-fat diet".
What it says now
It says it "lessened their weight gain on a western diet".
Why
In the paper (Li and colleagues, 2024), mice on a western diet given the bacterium in their water gained less weight and fat than the comparison mice. The paper calls the diet a western diet and does not give its fat share.
Li H, Zhang L, Li J, Wu Q, et al. Resistant starch intake facilitates weight loss in humans by reshaping the gut microbiota. Nat Metab. 2024;6(3):578-597.Read the paper
Found by
Our full re-read of the page against all its papers.
Hall et al. 20212 corrections · 1 page
November 20, 2026 · We got it wrong
A keto ward study page told the calorie gap without the diet it was measured against
What it said
Its key finding said the keto diet led to "689 more calories eaten a day, with no clear difference in hunger". A pull quote said "Seven hundred extra calories a day on the diet that was supposed to kill appetite". Its notes for AI systems said "Keto diet led to 689 extra calories/day" and "every single person ate more — 689 extra calories per day on average", "about three times higher on the high-carb diet, but energy intake went the other way, 689 calories per day lower" and "higher insulin, fewer calories". None of these lines said what the calories were measured against.
What it says now
The key finding now says the keto diet led to "689 more calories eaten a day than the plant-based, low-fat diet". The pull quote now says "Seven hundred more calories a day on the diet that was supposed to kill appetite than on the plant-based one". The notes for AI systems, which the page does not print, now name the diet each figure was measured against.
Why
In the trial (Hall et al., 2021), twenty adults ate both diets, two weeks each, as much as they liked. On the keto diet they ate 689 more calories a day, on average, than on the plant-based, low-fat diet. People lost weight on both diets, and the authors suggest that on keto, people ate about as much energy as they burned. Said without the other diet, the extra calories read as eating far more than the body needs, and the high-carb diet's lower intake as something that diet does on its own.
Hall KD, Guo J, Courville AB, et al. Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake. Nat Med. 2021;27:344-353.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A keto ward study page told AI systems that insulin rose in every participant
What it said
A note for AI systems said "The opposite happened in every participant — higher insulin, fewer calories".
What it says now
That note, which the page does not print, now says every participant ate fewer calories on the high-carb diet than on keto, and that insulin after the high-carb test meal ran higher on average.
Why
In the trial (Hall et al., 2021), every participant ate less on the plant-based, low-fat diet than on the keto diet. Insulin after a test meal was higher on the plant-based diet, but the paper gives that result only as the group's average, never person by person. The same words were the new wording of an earlier correction to another line of this page, about food (2026-10-16-keto-plant-based-study-food-volume-explanation-not-in-paper); that line has since been reworded.
Hall KD, Guo J, Courville AB, et al. Effect of a plant-based, low-fat diet versus an animal-based, ketogenic diet on ad libitum energy intake. Nat Med. 2021;27:344-353.Read the paper
Found by
Our full re-read of the page against all its papers.
Strozyk et al. 20229 corrections · 2 pages
November 20, 2026 · We got it wrong
A resistant starch study called cooling your rice a fine habit; the rice trial it cites found more low blood sugar after cooled rice in adults with type 1 diabetes on the same insulin dose
What it said
Its summary of what the findings mean said "Cooling your rice is a fine habit."
What it says now
That line is gone. The summary now ends "Counting on cooled rice for weight loss is a different claim entirely — one the evidence does not support."
Why
No paper this page rests on says cooling rice is a fine habit. The rice trial (Strozyk and colleagues, Nutrition & Diabetes, 2022) gave 32 adults with type 1 diabetes, all using insulin pumps, fresh rice and rice cooled for a day and reheated, each on the same insulin dose. After the cooled rice, 12 of the 32 had low blood sugar within three hours, against 3 after fresh rice. The authors write that cooled rice raises the risk of low blood sugar after the meal for people who do not lower their insulin dose. The correction of 20 November (2026-11-20-resistant-starch-study-told-a-rice-trials-smaller-blood-sugar-rise-as-a-plain-benefit-and-a-claim-called-the-rice-harmless) gave that low blood sugar beside every line that tells the smaller rise, but left this line, which called the habit fine.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim said cooled rice feeds good gut bacteria and gives a small gut health bonus; none of its studies measured that
What it said
The page said "The small amount of resistant starch in cooled rice feeds beneficial gut bacteria." and that a leftover serving "is giving you a small gut health bonus without trying".
What it says now
It says "What the studies here measured is blood sugar." and gives the rice trial's smaller rise in blood sugar with the low blood sugar beside it.
Why
None of the three studies on the page measured what cooled rice does to gut bacteria or gut health. The rice trial (Strozyk and colleagues, 2022) measured blood sugar after one meal in adults with type 1 diabetes. The gut-wall results in the weight-loss paper (Li and colleagues, 2024) were in mice given gut bacteria from people who took the supplement.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study gave the resistant starch and calories cooling saves per serving; the trial's figures were per 100 grams, and none of their studies measured calories
What it said
The claim said "A serving of cooled rice gives you roughly 4.4 extra grams" and that the calorie difference "per serving is roughly 20 calories". The study said "The extra resistant starch from cooling amounts to about 20 fewer calories per serving."
What it says now
The claim says "Cooling adds roughly 4.4 extra grams to every 100 grams of cooked rice." and that the trial's authors estimate cooling "leaves about 5 grams less carbohydrate to digest in every 100 grams of rice", which by FitChef's rough arithmetic "is at most about 20 calories". Both pages say "No study here measured the calories."
Why
In the rice trial (Strozyk and colleagues, Nutrition & Diabetes, 2022), cooling raised resistant starch from 7.52 to 11.96 grams in every 100 grams of cooked rice, and the authors estimate about 5 grams less digestible carbohydrate per 100 grams. Our arithmetic read the 4.4 grams as a serving's and used an energy value no paper gives. The pages now give the trial's basis, call the calories FitChef's estimate (5 grams of carbohydrate is about 20 calories, at most), and say the weight-loss supplement gave about nine times what cooling adds to 100 grams.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim said 10% or 15% fewer calories from cooling is not what the evidence shows; the rice trial's figures put the change at about a fifth of the carbohydrate
What it said
The page said "If you heard 10%, 15%, or 50% fewer calories, those numbers are not what the evidence shows."
What it says now
It says "If you heard that cooling halves the calories, no study here found that." and that the rice trial's own figures put the change at "about a fifth".
Why
No study on the page measured calories. In the rice trial (Strozyk and colleagues, 2022), 100 grams of the rice held 23 grams of carbohydrate, and the authors estimate cooling took about 5 grams of it out of digestion: about a fifth. That does not rule out a 10% or 15% drop in calories; it rules out only anything like half.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim said the rice trial had 15 participants; it had 32
What it said
Its evidence notes, which the page does not print, said "Small crossover trial (n=15) in T1D participants."
What it says now
Its evidence notes, which the page does not print, now say "Small randomized, single-blind crossover trial (n=32) in T1D participants."
Why
The rice trial (Strozyk and colleagues, Nutrition & Diabetes, 2022) included 32 adults with type 1 diabetes. The count of 15 came from our notes.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said reheating does not undo the resistant starch and gives the same benefit as cold rice; the trial never measured the rice cold
What it said
The claim said "meal-prepping rice and reheating it later gives you the same resistant starch benefit as eating it cold" and that "leftover rice that gets microwaved still has the same gut benefit as cold rice". The study said "Reheating does not undo the crystallization that formed in the fridge."
What it says now
They say rice "cooled for a day and briefly reheated in hot water still held more resistant starch than freshly cooked rice". The study adds "The trial measured resistant starch only after reheating, so it did not compare reheated rice with rice eaten cold."
Why
In the rice trial (Strozyk and colleagues, 2022), rice was cooled for 24 hours, reheated in hot water for three minutes and then measured; fresh rice was measured too. The cooled rice held more resistant starch, but no rice was measured cold or microwaved, and no gut effect was measured.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim said how much resistant starch forms depends on cooling time and temperature; its papers name other factors
What it said
Its answer said resistant starch formation "varies by starch type, amylose content, cooling time, and temperature".
What it says now
It says "as cooked starch cools, its amylose and amylopectin chains pack into tight structures that resist digestion" and that, in an earlier study, "repeated heating and cooling formed even more".
Why
The rice trial's authors (Strozyk and colleagues, 2022) describe resistant starch forming as cooked starch cools, when its amylose and amylopectin chains form tight double helices, and cite an earlier study in which repeated heating and cooling formed more. None of the papers tested cooling time or temperature.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study told a rice trial's smaller rise in blood sugar as a plain benefit, in lines that did not say the trial was in adults with type 1 diabetes or that more of them had low blood sugar; a cooled-rice claim's stored summary called the rice harmless
What it said
The resistant starch study page said "A crossover trial in 32 patients measured it directly" without saying they had type 1 diabetes. It said cooling rice "does reduce the blood sugar spike", that "the blood sugar data confirms a genuine benefit" and that "The blood sugar benefit is genuine", and never said the trial also found more low blood sugar. The cooled-rice claim's stored summary, which the page does not show, ended "Real, harmless, and not a weight-loss plan."
What it says now
The study page says "A trial in 32 adults with type 1 diabetes measured it directly", and wherever it tells the smaller rise in blood sugar it now names the people and gives the low blood sugar beside it: "Low blood sugar was also more common in the three hours after the cooled rice: 12 of the 32 had it, against 3 after fresh rice." The claim's stored summary, which the page does not show, now ends "Real, and not a weight-loss plan."
Why
The trial (Strozyk et al., Nutrition & Diabetes, 2022) gave 32 adults with type 1 diabetes, all using insulin pumps, fresh rice and rice that was cooled for a day and reheated. After the cooled rice their blood sugar rose less, but 12 of the 32 had low blood sugar within three hours, against 3 after the fresh rice, each on the same insulin dose for both meals. The authors write that cooled rice raises the risk of low blood sugar after the meal on a standard insulin dose. Our notes on the trial recorded who took part and the low blood sugar, but the study page's lines that told its result kept only the lower rise, and the claim's stored summary called the rice harmless.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said cooled rice cut the blood sugar spike by more than half; in the rice trial the spike was nearly a third lower
What it said
The resistant starch study page said cooled and reheated rice "cut the blood sugar spike by more than half" compared to freshly cooked rice.
What it says now
It says the rice "cut the blood sugar spike by nearly a third" compared to freshly cooked rice.
Why
In the trial (Strozyk et al., Nutrition & Diabetes, 2022), the highest rise in blood sugar after the meal was 2.7 mmol/L after the cooled rice against 3.9 after the fresh rice, nearly a third lower. What fell by more than half was the rise added up over the three hours, a different measure. Our notes held both numbers; the page gave the second one to the spike.
Strozyk S, Rogowicz-Frontczak A, Pilacinski S, LeThanh-Blicharz J, Koperska A, Zozulinska-Ziolkiewicz D. Influence of resistant starch resulting from the cooling of rice on postprandial glycemia in type 1 diabetes. Nutr Diabetes. 2022;12:21.Read the paper
Found by
Our full re-read of the page against all its papers.
A calories and fat loss claim's chart called 0.93 kg the extra weight lost on low-carb diets, without saying it was the difference from balanced diets
November 20, 2026 · We got it wrong
What it said
Its chart labelled the 0.93 kg "Extra weight lost on low-carb, after one to two years", and its descriptions for screen readers said "The extra weight lost on low-carb diets after one to two years".
What it says now
The label reads "More weight lost on low-carb vs balanced diets, after 1 to 2 years", and the descriptions for screen readers say "More weight lost on low-carb diets than on balanced diets after one to two years".
Why
In the review (Naude et al., 2022), the 0.93 kg is how much more weight was lost on low-carb diets than on balanced diets, in 14 trials of one to two years in people without type 2 diabetes. Both diets' groups lost weight: their average losses ranged from 2.46 to 13.1 kg on low-carb diets and from 1.7 to 11.6 kg on balanced ones, and the authors judged the difference probably too small to matter. Next to the 2 to 3 kg of body water the chart shows, "extra weight lost on low-carb" could be read as what a low-carb diet takes off beyond that water, or as all it takes off. When the chart was corrected on 17 November 2026 (record 2026-11-17-calories-claim-told-two-sets-of-trials-as-one-gap-shrinking-over-twelve-months), its label "Low-carb advantage after one year" became "Extra weight lost on low-carb, after one to two years", and the word that said it was a comparison was lost.
Naude CE, Brand A, Schoonees A, Nguyen KA, Chaplin M, Volmink J. Low-carbohydrate versus balanced-carbohydrate diets for reducing weight and cardiovascular risk. Cochrane Database Syst Rev. 2022;1(1):CD013334.Read the paper
Found by
Our full re-read of the page against all its papers.
Guarneiri et al. 202415 corrections · 3 pages
November 20, 2026 · We got it wrong
A protein digestion Short called the single-meal result settled and undisputed; the review found it on average, with results that varied widely
What it said
The page called the single-meal result "one of the most replicated findings in nutrition science" and said "That much is settled." Its recording said "this is exactly right" and "Nobody disputes this.".
What it says now
It says "Across 33 comparisons in 28 studies, higher-protein meals raised the calories burned after eating more than lower-protein meals did." and "Results varied widely from study to study, and the review rates the quality of this evidence as moderate."
Why
Guarneiri and colleagues (2024) pooled 33 single-meal comparisons from 28 studies. Higher-protein meals raised the energy burned after eating on average, but the results varied widely between studies, and the review rates the evidence as moderate. The rise was not clear in the studies of older or heavier volunteers, so it is not shown for every meal and every person.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein digestion Short said a result from four calorie-deficit comparisons undoes protein's advantage; the estimate was about as large as at maintenance, but uncertain
What it said
The page called the calorie-deficit result "the finding that undoes the entire strategy" and said "The evidence for the metabolic edge ran out at the precise moment a dieter would have used it." Its recording said "A fair-weather friend.", and its notes for AI systems said "The advantage holds only at eucaloric (maintenance) intake".
What it says now
It says "Only four comparisons measured total daily burn on a calorie-cut diet. The rise they found was about as large as at maintenance, but too uncertain to tell apart from chance."
Why
In the review by Guarneiri and colleagues (2024), the four comparisons made during a calorie deficit found a rise in total daily energy burned about as large as in the five at maintenance, but too uncertain to be sure of. A result that is not clear in four comparisons does not show that the advantage is gone.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein digestion Short said the effect is three times weaker in people with higher body weight; the review split studies, not people, and is unsure the gap is real
What it said
The page said the effect "is three times weaker in people with higher body weight", and its answer on body weight began "Yes — and the difference is large". Its answer on age began "It does. The per-meal thermic boost from protein is" and called the effect "strongest in young, lean individuals".
What it says now
It says "After single meals, it was clear in studies of leaner people but not in studies of heavier people, and the authors are unsure that gap is real." Its answers on body weight and age begin "It may".
Why
Guarneiri and colleagues (2024) split their single-meal studies at the middle value of their volunteers' average BMI, and at the middle of their average age. Higher-protein meals clearly raised the energy burned after eating more than lower-protein meals did in the studies of leaner and of younger volunteers, and not clearly in the other half. These are groups of studies, not people, and the two results are standardised scores whose ratio is not three times the calories. The authors say it is not clear whether the body weight difference is a true effect or comes from how the studies were run.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein digestion Short said the extra burn from protein meals never adds up over time and denied a lasting rise; the review found a modest lasting rise in daily and resting burn
What it said
The page said the advantage "does not accumulate over time" and "A per-meal fee that never compounded was never the real argument for protein." Its recording said "it does not compound" and "It never added up.", its chart labelled the longer-diet result "Stops adding up", and its notes for AI systems said the advantage "does not accumulate into a lasting metabolic elevation over weeks to months".
What it says now
It says "On those diets, the extra heat after a protein-rich meal was too small to tell apart from chance." and "Total daily burn and resting metabolism did end up modestly higher than on lower-protein diets of the same calories."
Why
Guarneiri and colleagues (2024) found that on higher-protein diets eaten for 4 days to a year, the extra energy burned after a meal was no longer clearly higher than on a lower-protein diet of the same calories. They raise the idea that the body adapts, and say it needs more study. They did not find that each meal kept its extra burn while the total failed to add up. On those longer diets, total daily burn and resting metabolism were modestly higher than on the lower-protein diets, on low-certainty evidence.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein digestion Short said protein's extra burn is enough to offset yearly weight gain but too small for fat loss; the review measured calories burned, not weight
What it said
The page said the rise was "Enough to offset the slow drift of 10 to 20 surplus calories a day that causes gradual annual weight gain" and "Not enough to drive intentional fat loss". Its answer for search tools ended "Too small to serve as a weight-loss lever.", and its recording said "Not enough to matter when you are actively cutting."
What it says now
It says "The authors judge a rise like this large enough to help limit that gain, if eating does not rise to match it." and "The review measured calories burned, not weight or fat lost."
Why
Guarneiri and colleagues (2024) write that 10 to 20 extra calories a day can explain the average yearly weight gain of US adults, and that the rise from a protein-rich diet is large enough to contribute to less weight gain if people do not eat more to make up for it. The review measured energy burned, not weight or fat loss, so it shows neither that the rise offsets weight gain nor that it is too small for fat loss. The page now gives FitChef's own reading as ours, in its body only.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein digestion Short said the rise in resting burn is likely because protein builds muscle; the review suggests more lean tissue may explain it
What it said
The page said the extra burn moves to resting metabolism "likely because higher protein supports more", and its recording said "perhaps because protein supports muscle, and muscle costs energy around the clock".
What it says now
It says "The authors suggest higher-protein diets may build more lean tissue over time, the main driver of resting burn."
Why
Guarneiri and colleagues (2024) suggest that over time higher-protein diets may result in greater fat-free mass, the main thing that sets resting energy expenditure, and say the shift to resting metabolism still needs confirming. Fat-free mass is more than muscle, and the review did not test the cause.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein digestion Short's recording told one trial's 3 percent as the listener's own daily burn
What it said
Its recording said "Three percent of your daily burn."
What it says now
Its recording now says "That was about three percent of that trial's daily burn."
Why
In the review by Guarneiri and colleagues (2024), the 72 calories a day, about 3 percent, was the difference in one trial the authors use as an example. It is not a share of any listener's own daily burn.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein Short's chart drew a size for the single-meal result that the review does not give, and called both results the same protein
What it said
Its chart filled the single-meal bar to two thirds of its track and was headed "SAME PROTEIN · TWO TIMESCALES".
What it says now
Its chart is headed "ONE MEAL · LONGER DIETS", and its source line says "the bars show whether there was a difference, not its size".
Why
Guarneiri and colleagues (2024) give the size of the single-meal result only as a standardised score, not as a share of anything a bar could show. The two rows pool different studies, 16 of single meals and 5 of longer diets, with different amounts of protein and different foods, so they were not the same protein at two timescales.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein Short's recording gave a cause for the fat-burning result that the review does not give
What it said
Its recording said "Because protein triggers a different metabolic response during digestion."
What it says now
Its recording now says "Here’s what the studies compared." and describes the two meals.
Why
Guarneiri and colleagues (2024) measured which fuel was burned after the meals and give no cause for the difference. They report it was much larger where mostly carbohydrate took protein's place in the lower-protein meal. The one explanation they offer is for the extra heat after meals, not for the fuel mix.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein Short gave a breathing measure of fuel use without saying the review trusts it less
What it said
The page said the respiratory exchange ratio "moved toward fat, pooled across 26 comparisons from 24 studies", with no word of how sure the review is of it.
What it says now
It says "But the review rates that measure’s evidence low, partly because studies that found nothing may be missing."
Why
Guarneiri and colleagues (2024) rate their single-meal results moderate, except the respiratory exchange ratio, which they rate low for its high heterogeneity and possible publication bias.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein Short told its single-meal fat-burning result as settled; the studies varied widely and the authors urge caution
What it said
The page said "The per-meal fat shift is real", and its recording said "This isn’t speculation." and "So your instinct was right. Protein genuinely changes which fuel your body burns after a meal."
What it says now
It says "The per-meal fat shift showed on average; on longer diets it did not.", and its answer for search tools says "That is the average of 18 tests in 16 studies, which varied a lot."
Why
Guarneiri and colleagues (2024) pooled studies whose results varied greatly, rate the evidence as moderate, and write that these results should be interpreted with caution. The page now gives the result on average, with the studies' disagreement beside it.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two protein Shorts called the review's longer diets weeks of eating; they ran from 4 days to a year
What it said
One page compared a single meal with "after weeks of protein-heavy eating". The other said "After weeks on the diet, the extra heat after a protein-rich meal was too small to tell apart from chance."
What it says now
The first says "on longer protein-heavy diets". The second says "But on higher-protein diets eaten for 4 days to a year, the picture changed."
Why
The longer-term studies in the review by Guarneiri and colleagues (2024) ran from 4 days to a year, several of them less than a week, and the review does not say how long the ones that measured fat burning or the energy burned after a meal lasted.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two protein Shorts tied a 52-study review's size to figures its 52 studies did not produce
What it said
One page said "In a 2024 meta-analysis of protein thermogenesis (52 studies, 1,232 participants), digestion is put at about 10% of total daily energy expenditure". The other said "In longer-term studies, a 52-study review found no effect".
What it says now
The first says "A 2024 review by Guarneiri and colleagues pooled 28 studies of single meals". The second says "In the 5 longer-term studies that measured it, it did not."
Why
In the review by Guarneiri and colleagues (2024), the 10 percent share of daily energy spent on digestion comes from its introduction, citing earlier work; its 52 studies compared higher- and lower-protein meals and diets and did not measure that share. Its longer-term result on fat burning after a meal came from 5 of the 52 studies.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
Two protein Shorts told results comparing higher- and lower-protein meals of the same calories as protein changing the reader's own body, one of them in its titles as FitChef's own finding
What it said
One page's title was "The Fat-Burning Shift That Follows a Protein Meal", its search title "A Protein Meal Raises Fat Burning. Longer Diets Showed No Effect.", its search description answered "Yes, after a single meal.", its share description said "After a protein meal, your body burns more fat for fuel.", its opening said "In the hours after a protein-rich meal, the body’s fuel mix tilts." and its chart said "Burns more fat". The other page's recording said "Protein does increase your total daily expenditure over time", and its chart said a single high-protein meal "Works".
What it says now
The first page's title asks "Does Protein Make Your Body Burn More Fat After a Meal?", and its opening says "A 2024 review found it did, on average, compared with a lower-protein meal of the same calories." The second page's chart now says "Higher burn" against "vs. a lower-protein meal".
Why
Guarneiri and colleagues (2024) compared meals and diets of the same calories with more and less protein, carbohydrate or fat taking protein's place in the lower-protein one. After the higher-protein meal, people burned more calories, and more fat and less carbohydrate, than after the lower-protein meal, on average, with results that varied widely; on longer diets, the higher-protein diet ended with a modestly higher daily burn than the lower-protein diet, on low-certainty evidence. These are differences between two meals or diets, not a switch the body makes after any protein meal, and no study tested a reader's own meals. In the 5 longer-term studies that measured it, there was no effect on fat burning after a meal. The recordings no longer tell the listener what their own body is doing, and the first page's titles and search lines now ask the question or name the review.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332. doi:10.1016/j.advnut.2024.100332.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A protein-digestion study page said a high-protein diet never raises resting calorie burn; the review found it a bit higher on longer diets
What it said
The study page said the promise that more protein permanently boosts resting metabolism was tested and failed, that "the permanent upgrade (chronic thermogenesis) was not statistically significant", that "Protein gives you a raise at every meal. It never promotes you to a higher base salary." and that "the boost works per meal, not permanently". It said a separate study "confirmed the same conclusion", and its answer to whether a high-protein diet permanently raises metabolism opened "The per-meal calorie-burning boost from protein did not become permanent in the pooled data." Its notes for AI systems said the boost "does not become permanent over weeks or months" and was "not a permanent metabolic upgrade".
What it says now
It says the review's answer "was more modest than the promise": on longer diets the heat after a meal was not clearly higher, but "daily and resting calorie burn were both a bit higher with more protein than with less", though the evidence is weak and the rise was clear only in studies shorter than about a month. Its summary line now reads "Match the calories, and the meal with more protein pays the bigger raise. Over longer diets, base salary ran a little higher too, though the evidence is weak."
Why
Guarneiri et al. (2024) pooled studies of higher- against lower-protein diets lasting 4 days to a year. On those diets the heat after a meal was not clearly higher, but total daily calorie burn and resting calorie burn were modestly higher. The review rates that evidence low or very low, and the rise was clear only in the studies shorter than about a month. The authors suggest the extra burn may shift from digestion to resting metabolism over time, and say this needs confirming. The page told the result for the heat after a meal as if the whole long-term rise had failed, though its own evidence section gave the rise. The words that tied a separate 2016 study to that conclusion are cut. An earlier correction (2026-11-16-protein-digestion-study-page-said-the-boost-is-not-permanent-and-gave-study-splits-to-individuals) fixed the same error in the page's key finding but not in these lines.
Guarneiri LL, Adams CG, Garcia-Jackson B, Koecher K, Wilcox ML, Maki KC. Effects of Varying Protein Amounts and Types on Diet-Induced Thermogenesis: A Systematic Review and Meta-Analysis. Adv Nutr. 2024;15(12):100332.Read the paper
Found by
Our full re-read of the page against all its papers.
A protein digestion Short gave protein's 20 to 30 percent digestion cost as fact; it is a generally accepted estimate that its review did not measure
November 20, 2026 · We got it wrong
What it said
The page said protein's calories "go toward digestion alone, compared to single digits for fat", its answer for search tools said "protein burns 20–30% of its calories during digestion", and its recording said "For every hundred calories of chicken, steak, or whey, your body spends twenty to thirty just digesting it."
What it says now
It says "A 2010 study by Barr and Wright quotes the generally accepted figures" and "The 2024 review this page rests on did not measure that range." The page now cites that study.
Why
The review behind the page, by Guarneiri and colleagues (2024), gives no percentage of protein's calories spent in digestion. Barr and Wright (2010) quote 20 to 30 percent for protein, 5 to 10 percent for carbohydrate and 0 to 3 percent for fat as the generally accepted figures, citing earlier work; they did not measure them. The page now says so, and its answer for search tools no longer gives the range as a fact.
Barr SB, Wright JC. Postprandial energy expenditure in whole-food and processed-food meals: implications for daily energy expenditure. Food Nutr Res. 2010;54:5144. doi:10.3402/fnr.v54i0.5144.Read the paper
Found by
Our full re-read of the page against all its papers.
A protein-digestion study page said the food record you keep undercounts by 11 to 47%; the review found food diaries off by as much as 41%
November 20, 2026 · We got it wrong
What it said
It said "The record you keep undercounts by 11 to 47%."
What it says now
It says "The food diary you keep can undercount by as much as 41%."
Why
Burrows et al. (2019) reviewed 59 studies that checked what people reported eating against doubly labeled water. Most methods under-estimated intake: by 11 to 41% for food records, which is what a food diary is, 1.3 to 47% for diet histories and 4.6 to 42% for food frequency questionnaires. No method ran from 11 to 47%, and not every study of food records found under-reporting. This is the same correction made on the review's own study page and the calories guide (2026-11-16-calorie-underreporting-study-page-overstated-a-review-and-a-mechanism).
Burrows TL, Ho YY, Rollo ME, Collins CE. Validity of Dietary Assessment Methods When Compared to the Method of Doubly Labeled Water: A Systematic Review in Adults. Front Endocrinol. 2019;10:850.Read the paper
Found by
Our legal review.
A cooled-rice claim and a resistant starch study traced the 50% calorie claim to a 2015 student presentation, with details no paper we hold reports
November 20, 2026 · We got it wrong
What it said
The claim said the 50% figure "traces to a 2015 presentation at a chemistry conference by an undergraduate researcher" that was "never peer-reviewed, never replicated". The study named the student and said "He never published the work in a peer-reviewed journal."
What it says now
They say the 50% figure "comes from none of the three studies here" (the claim) and "comes from none of the studies on this page" (the study), and set it against the rice trial's own figures.
Why
None of the papers these pages rest on (Li 2024, Strozyk 2022, Pugh 2023) mentions the presentation, its author, its coconut oil or its results. Those details came from our notes, not from a paper we hold, so they are removed.
Our full re-read of the page against all its papers.
Pugh et al. 20235 corrections · 2 pages
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said rice water holds less than a gram of resistant starch, or none, and works nothing like Ozempic; no paper they rest on measured it
What it said
The claim said the starchy water "likely contains less than 0.6 grams of resistant starch" and "They're not even the same category of mechanism." The study said rice water "does not contain resistant starch".
What it says now
They say "None of the studies here measured rice water." (the claim) and "None of the studies on this page tested it" (the study), and that in a 2023 review of resistant starch trials the gut hormone GLP-1 did not change.
Why
None of the papers these pages rest on measured rice water: those lines came from web searches, not from a pasted paper. The one related result we hold is from a 2023 review of resistant starch trials (Pugh and colleagues): across five trials that measured it, GLP-1, the hormone the trend's name points at, did not change.
Pugh JE, Cai M, Altieri N, Frost G. A comparison of the effects of resistant starch types on glycemic response in individuals with type 2 diabetes or prediabetes: A systematic review and meta-analysis. Front Nutr. 2023;10:1118229.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said a review of 36 trials found resistant starch consistently improved blood sugar in diabetes and prediabetes; it found that in type 2 diabetes, not prediabetes
What it said
The study said "A 2023 meta-analysis pooling 36 randomized trials confirmed that the same type of resistant starch consistently improved fasting blood sugar and insulin levels across populations with diabetes and prediabetes". The claim said "Across 36 clinical trials, resistant starch consistently improved markers of metabolic health in the populations studied."
What it says now
The claim says the review "looked at adults with type 2 diabetes or prediabetes" and that the type cooling makes "did not change blood sugar after a meal". The study says "Resistant starch lowered them in people with type 2 diabetes, but not in prediabetes."
Why
The review (Pugh and colleagues, Frontiers in Nutrition, 2023) included 36 trials and pooled 31. In its longer trials, the supplement's type of resistant starch lowered fasting blood sugar and insulin; split by condition, resistant starch lowered them in type 2 diabetes but not in prediabetes, and the type that cooling makes did not change blood sugar after a meal. Its authors say its conclusions should be drawn with caution.
Pugh JE, Cai M, Altieri N, Frost G. A comparison of the effects of resistant starch types on glycemic response in individuals with type 2 diabetes or prediabetes: A systematic review and meta-analysis. Front Nutr. 2023;10:1118229.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim rated the risk of publication bias low; the review it cites found signs of publication bias
What it said
Its evidence summary said "No obvious incentive for selective reporting." and rated the risk low.
What it says now
It says "Pugh 2023 detected publication bias in its acute postprandial glucose, GIP and HbA1c meta-analyses."
Why
The review (Pugh and colleagues, 2023) reports that publication bias was detected in three of its pooled analyses. The page's low rating did not match it.
Pugh JE, Cai M, Altieri N, Frost G. A comparison of the effects of resistant starch types on glycemic response in individuals with type 2 diabetes or prediabetes: A systematic review and meta-analysis. Front Nutr. 2023;10:1118229.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A cooled-rice claim and a resistant starch study said the supplement's starch resists digestion by its structure and acts differently in the gut; a review says high amylose, and that the types may act differently
What it said
The claim said the supplement "resists digestion because of its physical structure". The study said its type "exists naturally in raw, uncooked grains" and called the two types "Different chemistry in your gut."
What it says now
They say it "resists digestion because it is high in amylose" (the claim) and "The two types may act differently in your body." (the study).
Why
The 2023 review (Pugh and colleagues) describes this type of resistant starch as starch with a high share of amylose, found in high-amylose grains, and says the types may work through different mechanisms. None of the papers tested how the two types behave in the gut.
Pugh JE, Cai M, Altieri N, Frost G. A comparison of the effects of resistant starch types on glycemic response in individuals with type 2 diabetes or prediabetes: A systematic review and meta-analysis. Front Nutr. 2023;10:1118229.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study gave a review's doses as 10 to 40 grams and a Western diet's resistant starch as 3 to 8 grams a day
What it said
It said the review "found glycemic benefits at doses ranging from 10 to 40 grams" and that "a typical Western diet provides about 3 to 8 grams of resistant starch total per day".
What it says now
It says "In its longer trials that reported a dose, doses ran from 6 to 40 grams a day."
Why
The review (Pugh and colleagues, 2023) gives doses of 6 to 40 grams a day for its longer trials of two resistant starch types. None of the papers gives a Western diet's intake, or says the trial needed a supplement for that reason.
Pugh JE, Cai M, Altieri N, Frost G. A comparison of the effects of resistant starch types on glycemic response in individuals with type 2 diabetes or prediabetes: A systematic review and meta-analysis. Front Nutr. 2023;10:1118229.Read the paper
Found by
Our full re-read of the page against all its papers.
A cooled-rice claim said preparation, fat and meal make-up shift absorption two- to tenfold in controlled trials, naming none
November 20, 2026 · We got it wrong
What it said
The page said they had been "measured at two-to-tenfold magnitude in controlled trials".
What it says now
It points to the food science guide, which "looks at what else preparation method, fat pairing and meal composition change".
Why
The line named no trial, and none of the papers this page rests on gives that range. The food science guide is read against its own papers in its own audit.
Our full re-read of the page against all its papers.
A resistant starch study said the 1998 turmeric study was co-authored by the patent holder of its product and that four labs failed to repeat it, details no paper we hold supports
November 20, 2026 · We got it wrong
What it said
It said the 1998 study was "co-authored by the patent holder of the product it tested" and "Four independent labs have tried to reproduce it over 27 years. None succeeded."
What it says now
It says the claim "traces to a single 1998 study of eight men."
Why
These lines came from our notes, not from a paper we hold as pasted. They are removed; the turmeric pages are read against their own papers in their own audit.
Our full re-read of the page against all its papers.
Brown et al. 20042 corrections · 1 page
November 20, 2026 · We got it wrong
A resistant starch study said a salad trial used two dressings; it used three
What it said
It said researchers served the same salad "with two different dressings" and linked to "why fat-free dressing cancels your salad's nutrition".
What it says now
It says "with three different dressings" and links to "how fat-free dressing affects carotenoid absorption".
Why
In the salad trial (Brown and colleagues, American Journal of Clinical Nutrition, 2004), each person ate the same salad with three dressings, holding 0, 6 or 28 grams of canola oil. It measured how much of the salad's carotenoids were absorbed.
Brown MJ, Ferruzzi MG, Nguyen ML, Cooper DA, Eldridge AL, Schwartz SJ, White WS. Carotenoid bioavailability is higher from salads ingested with full-fat than with fat-reduced salad dressings as measured with electrochemical detection. Am J Clin Nutr. 2004;80(2):396-403.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A resistant starch study said the type of fat changes what reaches the blood; the salad trial behind it tested the amount of fat
What it said
It said "preparation method, fat type, and meal composition each independently" shift what reaches the bloodstream.
What it says now
It says "How food is made, the fat with it and the rest of the meal can change what reaches the bloodstream."
Why
The salad trial this line draws on (Brown and colleagues, 2004) compared dressings with different amounts of canola oil, not different fats. The food science guide is read against its own papers in its own audit.
Brown MJ, Ferruzzi MG, Nguyen ML, Cooper DA, Eldridge AL, Schwartz SJ, White WS. Carotenoid bioavailability is higher from salads ingested with full-fat than with fat-reduced salad dressings as measured with electrochemical detection. Am J Clin Nutr. 2004;80(2):396-403.Read the paper
Found by
Our full re-read of the page against all its papers.
Patterson et al. 201912 corrections · 1 page
November 20, 2026 · We got it wrong
A potato Short said a trial measured blood sugar for three hours; it measured it for two
What it said
The page said the trial measured "total blood sugar over three hours", and its recording said "Three hours of sampling, every fifteen to thirty minutes".
What it says now
It says "blood sugar over the full two hours, taken together, showed no clear difference", and its recording says the blood "was sampled over the next two hours".
Why
Patterson and colleagues (2019) took blood at 15, 30, 60 and 120 minutes after the potato was eaten, and their main measure was blood sugar over those 120 minutes. The three hours came from our own notes on the paper, which listed a 180-minute sample the paper does not have.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short said the body absorbed the same total glucose from hot and chilled potatoes; the trial measured blood sugar and found no clear difference
What it said
The page said "the total glucose the body absorbed over three hours came out the same", and its pull quote said how much total glucose "the body processed" did not change. Its recording said "The same. Hot or cold."
What it says now
It says "blood sugar over the full two hours, taken together, showed no clear difference".
Why
Patterson and colleagues (2019) measured blood sugar levels, not how much glucose the body absorbed or processed. Over the two hours, blood sugar did not differ clearly between the two potatoes, which in one trial of 30 women is not the same as equal. The authors suggest the chilled potato holds less digestible carbohydrate, so less glucose is available from it.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short called its blood sugar trial the only controlled trial; the trial itself reports an earlier study that found a lower total
What it said
The page called the trial "the only controlled trial" in its takeaway, its answer for search tools and its speakable summary, and said "every controlled trial available on this question was paid for, at least in part, by people who sell the product".
What it says now
It says "The trial's authors cite an earlier study that did find a lower total."
Why
Patterson and colleagues (2019) report an earlier study by Fernandes and colleagues, in which boiled red potatoes chilled for 12 to 24 hours lowered total blood sugar compared with hot ones. That study used another potato, men and women, and a group with a lower average BMI. So this trial was not the only one, and the answer on total blood sugar is mixed.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short gave a trial's insulin results without saying who took part, or that its chilled potatoes were baked and its hot ones boiled
What it said
The page said "When chilled potatoes replaced hot ones in a controlled trial, insulin response dropped" and "Insulin and GIP responses drop significantly after eating chilled potatoes", without saying who took part.
What it says now
It says "A trial gave 30 women with overweight or obesity a chilled baked potato on one visit and a hot boiled one on another."
Why
The trial by Patterson and colleagues (2019) studied 30 women with overweight or obesity, whose fasting blood sugar and insulin were above normal on average. The authors write that its findings cannot be extended to other groups. Its chilled potatoes were baked and its hot ones boiled, so the cooking method differed as well as the temperature.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short and its recipe cards applied a five-to-six-day fridge result to one night, to warm salads and to its recipes
What it said
Its recipe card said "Whether served warm or at room temperature, the potatoes carry more resistant starch than straight from the oven." and "The cooling step IS the resistant starch formation protocol this Short explains." The page said "Normal meal prep captures part of the effect", and its chart labelled the reheated potatoes "your meal prep".
What it says now
It says "Overnight in a home fridge is shorter, and neither study tested it." Its recipe cards say "A short cool like this one was not tested." and "Neither tested baby potatoes."
Why
Both studies kept their potatoes cold for five to six days before testing them: the trial by Patterson and colleagues (2019) for five days, the lab study by Raatz and colleagues (2016) for six. Neither tested one night in the fridge, a short cool to warm or room temperature, roasting or baby potatoes.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short called both studies' funders the potato industry and shortened one funder's name
What it said
The page said "both studies were funded at least in part by the potato industry — one by the Alliance for Potato Research" and that "every controlled trial available on this question was paid for, at least in part, by people who sell the product".
What it says now
It says "The trial was funded by the Alliance for Potato Research and Education." and "The lab study was funded by the United States Potato Board, the Northern Plains Potato Growers Association and the US Department of Agriculture."
Why
Patterson and colleagues (2019) thank the Alliance for Potato Research and Education for funding, and say the funders had no role in the study's design, data, analysis or writing; the paper does not say who the alliance is. Raatz and colleagues (2016) name the United States Potato Board, the Northern Plains Potato Growers Association and the US Department of Agriculture's Agricultural Research Service. This replaces the earlier correction 2026-10-20-cooling-potatoes-short-raatz-2016-funding, which had said the studies were funded "at least in part by the potato industry".
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short called its studies' findings confirmed, repeatable and real; each came from one study
What it said
The page said "Real, significant changes confirmed by blood draws." and "The findings hold up under the design." Its recording said "That part is measured, repeatable, and real" and "The findings are real."
What it says now
It says what each study measured and in whom; of the trial's insulin and GIP results it says "Both differences were clear."
Why
Each finding comes from one study. The trial by Patterson and colleagues (2019) studied 30 women and was not blinded, and its authors note that how its potatoes were grown and stored was not controlled. The study by Raatz and colleagues (2016) analysed three varieties from one region. Neither repeated its own result.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short said both of its studies measured resistant starch and tested reheating; only the lab study did
What it said
The page said the resistant starch formed at home "is likely less than what the studies measured", and its recording said "both studies that tested this".
What it says now
It says "The resistant starch forming in a typical kitchen is likely less than what the lab study measured."
Why
The trial by Patterson and colleagues (2019) measured blood markers, fullness and food intake; it did not measure the starch in its potatoes. Only the lab study by Raatz and colleagues (2016) measured resistant starch, and only it tested reheating.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short said potatoes do their best work warm; no source it rests on says so
What it said
The page linked to its potato fullness Short with the words "potatoes do their best work warm".
What it says now
It says that in the trial "the chilled potato left people no fuller", and links to that Short with the words "the most filling food" inside a question.
Why
The only source behind these pages that compares warm and cold potatoes for fullness is the trial by Patterson and colleagues (2019): it found no difference in how full people felt, and the chilled potato was rated less pleasant to eat an hour later. The study behind the linked Short tested boiled potatoes and did not compare temperatures.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short cited a trial under a title it does not have
What it said
Its citation gave the title "Postprandial glycemic and insulinemic response to resistant starch from potatoes".
What it says now
It gives the paper's own title, "Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin".
Why
The paper by Patterson and colleagues (2019) in Nutrients prints that title. The old title was composed, not copied from the paper.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short's section for researchers gave a spread for fasting blood sugar that the paper does not give
What it said
It said "fasting glucose 110.5 ± 14.9 mg/dL (elevated)".
What it says now
It says "fasting glucose 110.5 mg/dL (elevated)".
Why
Patterson and colleagues (2019) give the women's average fasting glucose as 110.5 mg/dL, with no spread. The 14.9 came from our own notes on the paper.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short said cooled potato starch packs tighter than before; neither study compares it with raw starch
What it said
The page said the starch chains "lock tight again as the temperature drops — tighter than before".
What it says now
It says they "lock tight again as the temperature drops, tight enough that digestive enzymes can no longer break them apart efficiently".
Why
Patterson and colleagues (2019) write that heat and water loosen the starch, and that chilling re-forms it into a packed structure. Neither study compares that structure with the starch of a raw potato.
Patterson MA, Fong JN, Maiya M, Kung S, Sarkissian A, Nashef N, Wang W. Chilled Potatoes Decrease Postprandial Glucose, Insulin, and Glucose-dependent Insulinotropic Peptide Compared to Boiled Potatoes in Females with Elevated Fasting Glucose and Insulin. Nutrients. 2019;11(9):2066. doi:10.3390/nu11092066. PMID: 31484331.Read the paper
Found by
Our full re-read of the page against all its papers.
Raatz et al. 20165 corrections · 2 pages
November 20, 2026 · We got it wrong
A potato Short gave resistant starch amounts from the authors' manuscript of a study; the published study's page we rest on gives only their order
What it said
The page said chilled potatoes held "roughly 40% more resistant starch" than hot ones and that reheating kept "about a third of what cooling added", and gave the amounts as 4.27, 3.45 and 3.00 grams per 100 grams, in its text, its answers and its chart.
What it says now
It says "Chilled potatoes held more resistant starch than the same potatoes served hot or reheated.", and its chart ranks the three without amounts.
Why
The amounts came from the authors' accepted manuscript of the study by Raatz and colleagues (2016), another version than the published study the page cites. The part of the published study we rest on, its journal page, gives the order but not the amounts: chilled potatoes had more resistant starch than hot or reheated ones. The page now gives the order alone. This replaces the earlier correction 2026-10-20-cooling-potatoes-short-raatz-2016-reheating-gain, which had set the chart's description to "about a third".
Raatz SK, Idso L, Johnson LK, Jackson MI, Combs GF Jr. Resistant starch analysis of commonly consumed potatoes: Content varies by cooking method and service temperature but not by variety. Food Chem. 2016;208:297-300. doi:10.1016/j.foodchem.2016.03.120.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short said reheating keeps most of the resistant starch intact; in the study, reheated potatoes had less than chilled ones
What it said
The page said "Reheating a cooled potato keeps most of the resistant starch intact."
What it says now
It says "In the lab study, reheated potatoes kept some of the extra resistant starch."
Why
In the study by Raatz and colleagues (2016), chilled potatoes had more resistant starch than either hot or reheated potatoes, and reheated ones came between the two. So reheating lost some of what cooling added; the study does not show it kept most of it.
Raatz SK, Idso L, Johnson LK, Jackson MI, Combs GF Jr. Resistant starch analysis of commonly consumed potatoes: Content varies by cooking method and service temperature but not by variety. Food Chem. 2016;208:297-300. doi:10.1016/j.foodchem.2016.03.120.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short gave the order of chilled, reheated and hot potatoes a significance and consistency its source does not state
What it said
Its section for researchers gave the order with "(p < 0.001)", and its notes for AI systems called it a "statistically significant hierarchy", "consistent across varieties".
What it says now
Its section for researchers gives the order in the study's own words, "regardless of cooking method, chilled > reheated > hot", with no p value for it.
Why
The study's journal page states that order but gives no p value for it. In the authors' accepted manuscript, which the old line was taken from, the p value below 0.001 compares chilled potatoes with hot or reheated ones, not reheated with hot, and reheated potatoes were not above hot ones for every variety and cooking method.
Raatz SK, Idso L, Johnson LK, Jackson MI, Combs GF Jr. Resistant starch analysis of commonly consumed potatoes: Content varies by cooking method and service temperature but not by variety. Food Chem. 2016;208:297-300. doi:10.1016/j.foodchem.2016.03.120.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
A potato Short said the type of potato makes no difference to resistant starch; the study tested three varieties from one region
What it said
The page said "The type of potato made no difference." and its answer said "The effect comes from the cooling process, not the potato."
What it says now
It says "The three varieties, all grown in one region, did not differ." and its answer names the variety Dark Red Norland, as the study does.
Why
Raatz and colleagues (2016) found no difference among three varieties grown in the Red River Valley of North Dakota and Minnesota: Dark Red Norland, Russet Burbank and Yukon Gold. The cooking method mattered too: baked potatoes had more resistant starch than boiled. Other varieties were not tested.
Raatz SK, Idso L, Johnson LK, Jackson MI, Combs GF Jr. Resistant starch analysis of commonly consumed potatoes: Content varies by cooking method and service temperature but not by variety. Food Chem. 2016;208:297-300. doi:10.1016/j.foodchem.2016.03.120.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short's link said hot potatoes cannot feed gut bacteria; the potato study behind it found resistant starch in all its potatoes
What it said
Its body linked to the Short on cooled potatoes with the words "cold potatoes feed gut bacteria that hot ones cannot".
What it says now
The link now says "chilled potatoes hold more resistant starch than hot ones".
Why
Raatz and colleagues (2016), the lab study behind the potatoes Short's starch results, measured resistant starch in potatoes baked or boiled and then served hot, chilled, or chilled and reheated. It found resistant starch in all the potatoes it tested, and more in chilled potatoes than in hot or reheated ones. It measured the starch, not gut bacteria. When this page was corrected on 20 November, its link words were checked against a 2024 review's general table of starch types, not against the potato study, and were kept.
Raatz SK, Idso L, Johnson LK, Jackson MI, Combs GF Jr. Resistant starch analysis of commonly consumed potatoes: Content varies by cooking method and service temperature but not by variety. Food Chem. 2016;208:297-300.Read the paper
Found by
One paper read again against every page that rests on it.
A potato Short said a review of glycemic index found a bigger answer than resistant starch; it found no clear difference in weight
November 20, 2026 · We got it wrong
What it said
The page said the glycemic index question "ran through fourteen trials and nearly two thousand people" and that "The resistant starch answer is small. The GI answer is not."
What it says now
It says "A review of 14 trials compared diets low and high in glycemic index. Weight did not clearly differ, in results from 1,770 people."
Why
Schwingshackl and Hoffmann (2013) pooled 14 trials comparing low and high glycemic index diets. Weight did not differ significantly between them, in the weight results of 1,770 people; the 14 trials enrolled 2,344 people in all.
Schwingshackl L, Hoffmann G. Long-term effects of low glycemic index/load vs. high glycemic index/load diets on parameters of obesity and obesity-associated risks: a systematic review and meta-analysis. Nutr Metab Cardiovasc Dis. 2013;23:699-706. doi:10.1016/j.numecd.2013.04.008.Read the paper
Found by
Our full re-read of the page against all its papers.
Baptista et al. 20248 corrections · 1 page
November 20, 2026 · We got it wrong
An oats Short called four kinds of gut bacteria the ones that break down resistant starch; its source reports only that they were more common
What it said
Its section for researchers said "Key RS-degrading gut bacteria include" Ruminococcus, Bifidobacterium, Faecalibacterium and Agathobacter.
What it says now
It says resistant starch "was associated with more Ruminococcus, Agathobacter, Faecalibacterium and Bifidobacterium", and with lower microbial diversity.
Why
Baptista and colleagues (2024) report pooled data linking resistant starch with more of these four kinds of bacteria, and with lower microbial diversity. They do not say these bacteria break the starch down.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short called the resistant starch in oats type 2, a type its source names for raw potatoes, green bananas and some legumes
What it said
Its section for researchers was headed "Resistant Starch (RS2) in Raw Oats", and an answer said resistant starch "(RS2) gets broken down when heat gelatinizes the starch granules".
What it says now
Its section for researchers says the review's authors state that uncooked oats are rich in resistant starch, and "They do not say which type of RS oats hold."
Why
In the review's table of types, starch trapped in whole or partly milled grains is type 1, and type 2 is found in raw potatoes, green bananas, some legumes and high-amylose corn. The review says uncooked oats are rich in resistant starch but does not say which type.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short's chart showed overnight and cooked oats side by side, credited to a review that never compared them
What it said
Its chart said overnight oats "Reaches your gut bacteria" and cooked oats are "Broken down before it gets there", credited "Starch granule integrity by preparation · Baptista et al. 2024".
What it says now
It shows starch in general, uncooked and cooked in water, credited "Starch in general, not a test of oats".
Why
Baptista and colleagues (2024) describe what heat and water do to starch granules in general. They never compare oats soaked overnight with cooked oats.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short graded its claims strong and moderate for comparisons no study had made
What it said
Its evidence notes for AI systems graded the claim that overnight oats keep more resistant starch "Strong — narrative review based on PubMed/Medline and Cochrane databases" and said "Mechanism well-established across multiple RS classifications". They graded the claim that the method decides the benefit "Moderate — individual mechanisms well-established".
What it says now
Its evidence notes for AI systems now grade the account of starch and heat moderate, and give the comparison of soaked and cooked oats as not tested.
Why
Neither review compared overnight-soaked with cooked oats. The review behind the starch claim says that uncooked oats are rich in resistant starch in one line citing an earlier paper, and Grundy and colleagues (2018) call the mechanisms not completely understood.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short's recording said any starchy food that cools keeps its resistant starch; its source says cooling forms new resistant starch, differently by food
What it said
Its recording said "Any starchy food that cools instead of staying hot preserves resistant starch your body treats as fiber."
What it says now
Its recording now says that cooked starch can form new resistant starch as it cools, and that how much forms depends on the food.
Why
Baptista and colleagues (2024) write that cooked starch partly re-forms resistant starch as it cools, and that the effect of cooling and storage on resistant starch appears to be specific to each food.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said the butyrate made from oat starch strengthens the gut barrier and lowers colon inflammation; its source does not say so
What it said
The page called butyrate "a compound that strengthens the gut barrier and lowers colon inflammation", said overnight oats feed gut bacteria "for colon health", and its recording said resistant starch "is what feeds the bacteria that keep your colon healthy".
What it says now
It says "In trials that gave adults 22 to 45 grams of resistant starch a day, one review found more" butyrate, and "Another review found no clear change."
Why
Baptista and colleagues (2024) report a review of nine trials that gave healthy adults 22 to 45 grams of resistant starch a day: butyrate rose. Another review found no clear change in butyrate. They put what people eat at 3 to 10 grams a day. The review says nothing about the gut barrier or colon inflammation. On inflammation markers in the blood, the reviews it reports give mixed results, and it says no consensus has been firmly established on inflammatory response.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said its line on oats drew on 14 meta-analyses; the line is one sentence citing another paper
What it said
The page said "the review synthesized evidence from PubMed, Medline and Cochrane databases across 14 meta-analyses, not a direct experiment" and "The direct comparison has not been run."
What it says now
It says "The 2024 review says uncooked oats are naturally rich in" resistant starch and "Neither review says anything about oats soaked overnight."
Why
Baptista and colleagues (2024) gathered 14 meta-analyses on resistant starch and health. Their line on uncooked oats sits in a separate section on food processing and cites an earlier paper. Neither review discusses oats soaked overnight, so neither can show whether such a comparison has been made.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said overnight oats skip heat entirely and keep a fiber the body can only get from oats left uncooked; its sources do not say so
What it said
The page said "Overnight oats skip the heat entirely. The crystalline structure stays locked." and called resistant starch "a type of fiber your body can only access when you do not cook the oats".
What it says now
It says "Overnight oats are not cooked in your kitchen, but they are not raw either." and "What soaking does to the starch in rolled oats, neither review measured."
Why
Grundy and colleagues (2018) note that oats are steamed as part of milling, and that oat flakes are often partly precooked before they are dried. Baptista and colleagues (2024) say uncooked oats are naturally rich in resistant starch and that cooking starch in water makes it more digestible, but some forms again as cooked starch cools, and the starch trapped in whole grains does not break during cooking. Neither review measured resistant starch in oats soaked overnight or cooked as porridge.
Baptista NT, Dessalles R, Illner AK, Ville P, Ribet L, Anton PM, Durand-Dubief M. Harnessing the power of resistant starch: a narrative review of its health impact and processing challenges. Front Nutr. 2024;11:1369950. doi:10.3389/fnut.2024.1369950. PMID: 38571748.Read the paper
Found by
Our full re-read of the page against all its papers.
Grundy et al. 20189 corrections · 1 page
November 20, 2026 · We got it wrong
An oats Short credited its 6% and 9% cholesterol drops to a review of more than 40 trials; they came from two trials of cooked oat flakes
What it said
The page said "A 2018 review of more than 40 human trials on oat processing found that cooked oat flakes reduced total cholesterol by roughly" 6%, and its answer for search tools said oatmeal "reduces LDL cholesterol by roughly 9% (Grundy et al. 2018, Food & Function, 40+ human trials reviewed)".
What it says now
It says "In two trials of oat flakes cooked with milk or water, total cholesterol fell by about" 6% and LDL by about 9%.
Why
The 6% and 9% come from two trials, reported by Grundy and colleagues (2018), in which oat flakes were cooked with milk or water. Their table summarises about 40 human studies of many oat foods, not more than 40 trials of cooked oats.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short gave a molecular weight range oat fiber must stay above to lower cholesterol; its source gives no such range
What it said
Its section for researchers said "Beta-glucan molecular weight should remain above 530–1200 kDa for cholesterol-lowering activity."
What it says now
It says "Kim et al. suggested at least 1200 kDa, Wolever et al. (2010) found cereals of 530 kDa effective at 3 g/day".
Why
Grundy and colleagues (2018) quote one group's suggestion of at least 1200 kDa, then a study in which cereals of 530 kDa lowered cholesterol at 3 grams a day while 210 kDa halved the effect; they add that viscosity depends on more than molecular weight.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short gave the wrong PubMed number for its oat processing review
What it said
Its list of sources gave the review's PubMed number as "29372218".
What it says now
Its source list in the structured data now gives the review's PubMed number as 29431835.
Why
The review as pasted from PubMed Central prints its PubMed number as 29431835.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said 40 to 60 grams of oat flakes give the 3 grams of beta-glucan a day that health claims name; its source gives no such amount
What it said
Its answer said the claims were for "3 grams per day" and "That is roughly one standard bowl of oatmeal made from 40-60 grams of oat flakes."
What it says now
It says the claims "say a food should give at least" 3 grams of beta-glucan a day, and that in two trials "people ate 70 to 100 grams of oat flakes a day, cooked with milk or water".
Why
Grundy and colleagues (2018) report that the health claims ask for at least 3 grams of beta-glucan a day. They give no beta-glucan content for oat flakes, so the page's bowl size had no source. In the two trials of cooked oat flakes, people ate 70 and 100 grams of oats a day.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said cooking unlocks the fiber in oats that lowers cholesterol, and that oats must be cooked, not soaked; its source does not say so
What it said
The page called beta-glucan "a soluble fiber that heat unlocks rather than destroys" and said this "which is how cooking your oats actively" lowers LDL cholesterol. Its answer on how much oatmeal to eat said the oats need to be cooked, "not just soaked, because heat dissolves beta-glucan into the viscous gel that traps bile salts".
What it says now
It says "The review says heating oats in water usually frees more beta-glucan." and "But in the authors' own lab test, oats mixed with water and slowly brought to the boil released less."
Why
Grundy and colleagues (2018) write that heating oats in water generally frees more beta-glucan, but in their own lab test, oats mixed with water and gradually brought to the boil released less. They found no clear pattern in the cholesterol results by food form, and they say the ways beta-glucan lowers cholesterol are not completely understood. Nothing in their review says soaked oats fail to lower cholesterol.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said soaking oats serves the gut and cooking them serves cholesterol; neither review it cites compared the two
What it said
The page said "The preparation method decides which health benefit the same oats deliver" and "Soak overnight for gut health, cook for cholesterol support." Its recording said "You are choosing between gut health and cholesterol support."
What it says now
It says "Neither review compared overnight oats with cooked oats." and "What the method changes for cholesterol, neither review tested."
Why
The page rests on two reviews. Baptista and colleagues (2024) write about resistant starch and how food processing changes it, and never mention oats soaked overnight. Grundy and colleagues (2018) review how processing changes the effect of oats on cholesterol: they found no clear pattern by food form, solid, liquid or porridge, and in their own lab test, oats brought slowly to the boil in water released less of the fiber beta-glucan, not more. Neither review compared soaked and cooked oats, so the page's split of benefits was its own idea, told as a finding.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short said porridge cooked on the stove was the minimally processed form that worked better; its source does not say so
What it said
The page said "Minimally processed oats, the porridge you make on the stove, were more effective than highly processed alternatives." and its answer said "Highly processed oat products are less effective than" minimally processed oat flakes cooked on the stove.
What it says now
It says the review "says oat foods that keep more of the oat's own structure often seemed to work better than pure beta-glucan added to food" and "Its authors say this still needs more study."
Why
Grundy and colleagues (2018) suggest that oat foods in which some of the plant tissue stays intact often appear to lower cholesterol better than purified beta-glucan added as an ingredient, one of three ideas they say need further study. By minimally processed they mean the oat tissue stays mostly intact; they do not name porridge as that form.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short told how oat fiber lowers cholesterol as settled; its source calls it not completely understood
What it said
The page said "The mechanisms are well-established." and that the fiber's gel "traps bile salts and prevents them from being reabsorbed".
What it says now
It says the review "says it is thought to lower cholesterol in several ways" and that "the exact ways are not fully understood".
Why
Grundy and colleagues (2018) write that the precise mechanisms are not completely understood, that beta-glucan is thought to lower cholesterol through several processes, and that its effect on bile salts still remains unclear.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
November 20, 2026 · We got it wrong
An oats Short's list of oat flake trials left out the one in which total cholesterol did not fall significantly
What it said
Its section for researchers listed five trials under "Human Trial Data (Cooked Oat Flakes, from Grundy 2018 Table 1)".
What it says now
It lists six, adding "Bindu 2013: 35g oat flakes → TC not significant, LDL-C −8%."
Why
The review's table lists six oat flake studies in which the flakes were eaten semi-solid. In one, Bindu 2013, total cholesterol did not change significantly and LDL fell by 8%. The table does not say the flakes were cooked in every study.
Grundy MM, Fardet A, Tosh SM, Rich GT, Wilde PJ. Processing of oat: the impact on oat's cholesterol lowering effect. Food Funct. 2018;9(3):1328-1343. doi:10.1039/c7fo02006f. PMID: 29431835.Read the paper
Found by
Our full re-read of the page against all its papers.
A sleep and exercise study page and its claim still told AI systems an age band and training levels for a review's people that our copy of the review does not give
November 20, 2026 · We got it wrong
What it said
The tags FitChef gives AI systems and apps for the study page called the review's participants "young-adults", "trained", "untrained" and "mixed-training"; the claim page's called them "mixed-training-status".
What it says now
The study page's tags now say only that the participants were adults, men and women, with no known medical conditions. The claim page's say adults with no known medical conditions, mostly men. Neither gives a training level.
Why
The review (Craven et al., 2022) took adults aged 18 and over, men and women, with no known medical conditions; the text we hold gives no ages beyond that. Its authors list training status among the details they collected about the participants and write that they could not explore its influence; they set out the participants' details in an online appendix, which we do not hold. Two earlier corrections mended these points in the study page's own words: 2026-10-17-craven-2022-study-page-population-and-figures-misreported (the participants' ages) and 2026-11-17-sleep-and-workout-study-page-said-a-reviews-participants-had-mixed-training-levels (their training levels). These tags, made from the pages' early notes, were not read then, because the audit's view of a page did not show them.
Craven J, McCartney D, Desbrow B, Sabapathy S, Bellinger P, Roberts L, Irwin C. Effects of Acute Sleep Loss on Physical Performance: A Systematic and Meta-Analytical Review. Sports Med. 2022;52(11):2669-2690.Read the paper
Found by
Our own standing checks.
A recipe card still showed a fiber Short's old title, which told a comparison of short and long trials as a personal timeline
November 20, 2026 · We got it wrong
What it said
Its research card linked the Short on fiber and weight loss under the title "Fiber for Weight Loss Takes 8 Weeks, Not More Grams". That Short showed this title in search results until 18 November 2026.
What it says now
The card shows the Short's current title, "How Much Fiber for Weight Loss? What 62 Trials Found".
Why
Jovanovski and colleagues (2020) report that trials shorter than eight weeks found a 0.08 kilogram reduction and trials longer than eight weeks 0.82 kilograms, suggesting that taking viscous fiber over a longer period may lead to lower body weight. That compares groups of trials, with different fibers, doses and people; it does not follow the same people week by week, and the shortest trials ran four weeks. The review also found no evidence that the daily amount of fiber changed the result, which is not the same as showing that more grams do not help. The Short was corrected on 18 November 2026 (record 2026-11-18-fiber-dose-short-told-short-and-long-trials-as-a-personal-timeline), and that record names the pages that showed its title; this card showed the title the Short carried in search results instead, so that record does not name this recipe.
Jovanovski E, Mazhar N, Komishon A, Khayyat R, Li D, Blanco Mejia S, Khan T, Jenkins AL, Smircic-Duvnjak L, Sievenpiper JL, Vuksan V. Can dietary viscous fiber affect body weight independently of an energy-restrictive diet? A systematic review and meta-analysis of randomized controlled trials. Am J Clin Nutr. 2020;111:471-485.Read the paper