There Are Two Kinds of “No Evidence”, and Only One of Them Means the Claim Is Wrong
You have read the sentence a hundred times. There is no evidence for that.
Did it mean nobody checked, or that somebody checked and found nothing?
Those are opposite situations, and the same five words cover both. One says a claim was put to the test and failed. The other says the test was never run — so the claim might be perfectly true, and nobody on earth can currently tell you.
We have been keeping track. Below are the ones we hit while checking claims one at a time, sorted into which kind they are.
Say this part out loud first
An untested claim is not a disproven claim. Nothing in the first list below has been shown to be false by anything we found. We looked for the test and did not find one, which is a different and much weaker thing than showing that nobody ran it.
Some rows below sit on verdicts we rate Unsupported. Since 7 October 2026 that rating means we looked and the evidence does not back the claim: either it was tested and did not hold up, or nobody has tested it and nothing that was tested supports it. Each verdict says which. A row here is the second kind, and it is still not a disproven claim.
We are labouring this because the drift is so easy. A site that collects gaps starts treating every gap as a debunk, and within a year it is the site that says nothing works. An absence is a description of what a search of the literature turned up, not a verdict on the claim.
What an absence does tell you is that anybody stating the claim confidently — in either direction — is going beyond what is known. That includes us, and on 30 September 2026 we found out how much: an audit of this page turned up a study, or part of one, behind twelve of the eighteen entries the first list then had.
Kind one: nobody ran the study, as far as we can find (11 of them)
Each line names the specific study we looked for and did not find, and says where we looked and when. PubMed and Europe PMC are indexes of published medical research; Crossref is a wider index of journal articles, reaching back further; the two trial registries are ClinicalTrials.gov and ISRCTN, where trials are listed before they start. That precision is the point — “no evidence” is exactly the vague phrase we are trying to replace.
| Lymphatic drainage massage, for detox and bloating The reviews behind our verdict pooled trials in women treated for breast cancer, most of them with lymphedema. For flushing out toxins or easing bloating in healthy adults, we found no trial that measured a toxin or bloating (PubMed and two trial registries, 30 September 2026). The nearest things are small trials in healthy people after hard exercise, which measured lactate, a by-product of effort that builds up in the blood, or muscle enzymes. We count seven such experiments, reported in eight papers. Five, in 14, 18, 80, 16 and 36 people, reported lower levels after drainage than after rest [1][2][3][36][37]; two, of 30 people each, found no significant difference [4][38]; and the one that also tried an ordinary massage found the drainage no better than it [36]. And in one trial of some fifty students with constipation, drainage was compared with two other treatments and nobody went untreated [5]. | unsupported |
| Oil pulling, for drawing out toxins The trials behind our verdict measure plaque, gum inflammation and bacteria, and between them breath, dry mouth, saliva chemistry and tooth decay as well. For drawing out toxins, a claim the videos we sampled make, we found no study that measured a toxin in anyone before and after oil pulling (PubMed and two trial registries, 30 September 2026). The other claim, whiter teeth, has been measured several times and the answers do not agree, so it is on neither list: a four-day trial found three oils changed tooth colour about as much as water did [6], and a 15-day trial of one maker’s rinse reported teeth slightly lighter than on a placebo, judged by eye [7]. The note at the foot of this page has the rest. | preliminary |
| The best-selling testosterone boosters, as sold For a review printed in 2021, researchers took the 16 best-selling testosterone supplements and the 16 best-selling erection supplements, pulled out all 37 ingredients, and found 105 trials of the ingredients. Trials of those 32 products as sold: their search of PubMed found none [8]. That is the review’s search, made before the paper was submitted in January 2020, and we could not repeat it: the products are named in a part of the paper we could not read. A few other finished products have been trialled, and the record is thin. | unsupported |
| Beef tallow on skin A 2024 scoping review found tallow creams moisturising in small tests, and still concluded that what is needed is research on tallow as a cosmetic product in humans [9]. We found no trial of rendered beef tallow, as sold, on human skin (PubMed, Europe PMC, Crossref, OpenAlex and two trial registries, between 30 September and 7 October 2026). The nearest things are three small studies from one Polish research group of laboratory creams made from mutton tallow blended with plant oils: 83 and 78 people tried them and rated them [10] [40] (the 78 were women with dry, sensitive skin and no signs of eczema, whose skin moisture was metered for 90 minutes), and 10 volunteers had their skin moisture metered for under three hours [41]. None set the cream against one you could buy. | unsupported |
| Testosterone screening Stated plainly in the 2018 clinical guideline: no clinical trials have assessed whether screening strategies actually help [11]. We found none published or registered since (PubMed and two trial registries, 30 September 2026). | unsupported |
| A home test for loose skin We found no validated way to tell loose skin from fat at home: not a pinch, not a threshold (PubMed, 30 September 2026). The questionnaires that have been validated measure how much excess skin bothers someone after weight loss. A Dutch study of what those answers actually track published its plan in 2025 [12]; we found no results from it yet. | unsupported |
| Food combining, for bloating Promoted continuously since 1929. For the claim that separating foods eases bloating, we found no trial that gave people separated or mixed meals and counted the symptoms (PubMed, two trial registries and Crossref, 30 September 2026). The mechanism behind the claim, that protein and starch get in each other’s way in the stomach, was tested in 1934 and again in 1936, and is in the second list. | unsupported |
| Sea moss and heavy metals Sold as removing heavy metals; the US National Institutes of Health’s reference record on it says it contains “some toxic heavy metals such as arsenic, lead, cadmium and mercury”. Whether these cancel out in a body, we found no study that measured (PubMed and two trial registries, 30 September 2026). Other seaweeds have been fed to volunteers: in a 2024 study, 11 people ate nori, kombu and wakame, and the arsenic in their urine went up afterwards [15]. A Norwegian study of 49 regular seaweed eaters registered cadmium and arsenic in urine among the things it would measure; the paper it published in 2026 reports iodine and thyroid results only [16]. One thing a web search will show you: a seaweed seller’s blog describes a 2022 study in which 24 adults took a sea moss supplement and had their cadmium tracked for three weeks, and names the journal. That journal is indexed in full in PubMed, and neither PubMed nor Crossref holds such a paper [17]. | unsupported |
| Intermittent fasting, for hard outcomes A Cochrane review, searching to December 2019, found no trial that reported a heart attack, a stroke or a death [18]. One has since. A 2024 pilot randomised 48 people who had just had a heart attack and analysed 42 of them over six months: one of the 20 on a regular diet died and none of the 22 who fasted, and one in each group went back to hospital with another coronary event [19]. Its authors call it a feasibility study and ask for a trial large enough to answer the question. We found none, published or registered (PubMed and two trial registries, 30 September 2026). | preliminary |
| Today’s apple cider vinegar gummies In 2005 researchers tested eight ACV tablets and could not confirm they contained apple cider vinegar. In 2020 a testing company measured today’s gummies and found a small fraction of the acetic acid in a tablespoon of vinegar. Whether a gummy does anything for weight or blood sugar, we found no trial, published or registered (PubMed and two trial registries, 30 September 2026). Tablets have been tried: in twelve healthy young adults in 2020, four vinegar tablets swallowed with a starchy meal left the rise in blood sugar where water left it, while liquid vinegar cut it by about a third [20]. | preliminary |
| Most of the Blue Zones A 2025 review found ten places that study authors had called Blue Zones [21]. Three are well characterised, and two of those are fading. Loma Linda, one of the brand’s original five, has been studied as a Blue Zone more than once: in 2024 researchers at the university hospital there compared 1,032 hip-fracture patients from inside and outside the zone, and reported that the zone’s patients were on average twelve years older when the hip broke and that fewer of them had died two years after surgery [22]. That is a study of one hospital’s patients, not of the town. What we did not find is the study that would show the town is a Blue Zone: a check that the people who live there outlive their neighbours (PubMed and Europe PMC, 30 September 2026). The review found no such study either, and a 2025 paper co-written by the demographer whose 2004 paper named the first Blue Zone lists Loma Linda among the places visited and not validated [23]. | preliminary |
Kind two: somebody ran it, and it came back empty (8 of them)
Here is the contrast. These claims got a test, and it came back empty. The tests are not equally strong. Some are randomised trials of dozens or hundreds of people; others are three volunteers, five volunteers, fifty hospital patients who were their own comparison, or before-and-after studies with no comparison group. Each line says which.
| Creatine and hair loss Forty-five men recruited, thirty-nine randomised, thirty-eight completed, blinded, twelve weeks. They measured DHT, the DHT-to-testosterone ratio, hair density, follicular units and hair thickness. The result: nothing. | unsupported |
| The post-workout testosterone spike About as direct a test as you can run: identical arm curls, one arm followed by a deliberate hormone surge and the other not. Strength rose equally in both. In the paper’s words, no effect of hormone elevation. Two small experiments elsewhere saw a strength edge; neither showed the surge builds the muscle. | unsupported |
| Food combining, for weight Fifty-four people, calories and macronutrients matched. Separated group lost 13.7 lb (6.2 kg), mixed group lost 16.5 lb (7.5 kg). No significant difference, and the numbers ran the wrong way. | unsupported |
| Berberine as “nature’s Ozempic” The whole pitch rests on raising GLP-1. In rats, it does. In people, a double-blind trial measured it in 300 adults with raised blood sugar over 16 weeks, and berberine moved it no more than a placebo did. | unsupported |
| Lymphatic drainage massage, for weight Two small studies weighed people before and after a course of it. In 15 healthy women with cellulite, after fourteen weekly sessions, body-mass index (weight relative to height) and body-fat percentage showed no significant fall [26]. In 30 adults picked because their bodies handled sugar poorly, 16 of them overweight or obese, body-mass index, waist-to-hip ratio and visceral fat were the same a month after ten sessions [27]. Neither study had a comparison group. A third, in 28 women office workers in Korea, compared a slimming cosmetic with and without abdominal drainage, and nothing, for four weeks: body-mass index and waist-to-hip ratio changed within both cosmetic groups and the authors judged the drainage version better, but the English abstract does not say how the groups were formed or what happened to the women given nothing [39]. | unsupported |
| Castor oil packs, for crossing the skin The mechanism needs castor oil to cross skin. A 2005 study measured it in three volunteers: after they swallowed castor oil, its breakdown products rose in their urine; after an abdominal hot pack, they did not. The authors concluded that it is either not well absorbed through skin or is broken down some other way. Three people is very few, and the study was paid for in part by the organisation that publishes instructions for the packs [31]. | unsupported |
| Food combining, for digestion The mechanism, that protein and starch eaten together get in each other’s way in the stomach, was tested in people at least twice in the 1930s. In 1934 fifty hospital patients ate hamburger steak one morning and steak with mashed potato the next: their stomachs emptied in three hours fifty-one minutes and three hours fifty-four [32]. In 1936 five people swallowed drinks of protein, of starch and of both while researchers took more than 900 samples of their stomach contents: the mixture did not interfere with digestion in the stomach, and the starch was broken down further with protein present [33]. | unsupported |
| Cold plunges, for a calorie deficit The calorie figure comes from studies scaling hours of cold to a full day. A cold bath itself has been measured. In 2025, 15 adults sat chest-deep in 61 °F (16 °C) water for half an hour, ten times the length of a three-minute plunge. On average they burned about 21 more calories than in warm water (89 kJ), and ate about 230 more (966 kJ) at the meal they were given half an hour later [34]. The calorie figures are our conversions, and that is one experiment and one meal. For three minutes itself we found no measurement (PubMed, 30 September 2026). | unsupported |
One claim manages to be both at once
Food combining — the idea that you should not eat protein and starch in the same meal — has been promoted continuously since 1929, and it makes two promises.
The weight promise was tested. Fifty-four people, calories and macronutrients matched, which is exactly the right way to isolate the separating. The separated group lost 13.7 lb (6.2 kg); the mixed group lost 16.5 lb (7.5 kg). No significant difference, and the numbers ran the wrong way.
The mechanism behind the digestion promise was tested too, in 1934 and again in 1936, and it failed both times. Fifty hospital patients ate steak one morning and steak with mashed potato the next, and their stomachs took three minutes longer over the second meal [32]. Then five people swallowed drinks of protein, of starch and of both while researchers sampled their stomach contents more than 900 times, and the mixture did not interfere with digestion in the stomach [33]. Until 30 September 2026 this page said the digestion promise had never been tested at all. Neither paper is indexed in PubMed, so a search there comes back empty.
What we have not found is a trial of the promise as people feel it: one that gave people separated or mixed meals and counted the bloating (PubMed, two trial registries and Crossref, 30 September 2026). So the same diet still sits in both lists. If you only ever heard “there is no evidence for food combining”, you would not know that its weight promise and its digestion mechanism were both measured and both failed, or that the bloating is the part we cannot find a measurement of.
So which kind should worry you more?
The untested kind, and it is not close.
A claim that was tested and failed is a solved problem. It is unpleasant for whoever was selling it, but the information exists and you can act on it.
An untested claim is worse for you specifically, because it cannot be resolved by reading harder. If the test was never run there is no paper to find, and everyone confidently discussing it — the people selling it and the people mocking it — is working from the same nothing. The trouble is telling “never run” from “not found”. We could not: behind twelve of the eighteen claims this list began with there was a study, or part of one, that the entry said did not exist.
And notice which claims end up in the first list. Not obscure ones. The best-selling testosterone boosters, lymphatic drainage for detox, apple cider vinegar gummies — large, busy markets where neither we nor the reviewers we cite could find a test of the specific question a buyer would ask.
Our read on why: the study that settles a claim is only worth funding to somebody who does not already have your money. A market can run for decades on a plausible mechanism, and the people inside it have little reason to spend on the experiment that could end it. That is a guess about incentives, not a finding, and what turned up on 30 September 2026 cuts both ways. The creatine gummies were bought and tested by a supplement maker checking other companies’ products and by a firm promoting its own ratings, which fits the guess. The castor oil study was paid for in part by the organisation that promotes the packs, which does not.
How to tell which one you are reading
This is the part worth taking away, because it works on claims we have never touched.
- Ask what the study would have looked like. If someone tested this properly, who was in it, what got measured, how long? If a page cannot tell you that, it is probably reporting an absence without knowing it.
- Watch for ingredient evidence standing in for product evidence. “Studies show ashwagandha…” is a claim about a compound. Whether the capsule in your hand contains a useful amount of it is a separate question, and usually an unasked one.
- Watch for the population swap. Real trials, run in patients after cancer surgery or in rats, quietly re-pointed at a healthy adult. The evidence is genuine; it is just about somebody else.
- “Not proven” and “disproven” are not synonyms — and anyone using them interchangeably is telling you something about their argument rather than the research.
What happens to this page next
It grows, and we would rather it grew in one direction than the other.
Every absence here is recorded as an absence in our searching, not a proof that nothing exists. If a study turns up, that entry moves to the second list or off the page, and we will say which. A register like this is only worth anything if entries can leave it. On 30 September 2026 nine did, in whole or in part: two moved to the second list, four left the page, two were split between the two lists, and one lost the half of its claim that had been measured after all, with answers that do not agree. The note at the foot of this page says which. On 1 October 2026 one more left the page, walking after meals, for weight, and the newest note says why.
The one we most wanted to see filled was creatine gummies, and it had been filled before we asked. We called it the cheapest unanswered question on this page, and said it had stayed unanswered because answering it served nobody who profits from the argument. A supplement company had bought twelve brands, assayed them in its own laboratory and reported in February 2024 that six fell short of the creatine on the label, several of them with “significant amounts” of creatinine, which is what creatine breaks down into. But the company’s own reading was that even the creatine and the creatinine together fell short of the labels, so those products had apparently not been given enough creatine to begin with [24]. A supplement-rating firm sent six brands to an outside laboratory and reported in June 2025 that four failed, two of them with no creatine detected [25]. Neither report is peer-reviewed and neither firm is disinterested: the first sells supplements of its own, the second promotes its own rating system, and one gummy maker called the first result inaccurate. The other brands passed. A result split like that is not a claim that came back empty, so the entry is now on neither list. What we have still not found is the version with nobody’s product attached: an analysis published in a journal (PubMed and Europe PMC, 30 September 2026).
The four ratings above are the ones on every verdict here, and we wrote down how we read a study so you can see what earns each. Two companion pages measure different things from the same corpus: what health advice actually reaches people, and who paid for the study.
Nobody pays us to reach a conclusion, and there is nothing for sale on this page.
An independent editorial review of our creatine hair-loss verdict on 18 September 2026 found an error this page repeated. Our entry for that trial said forty-five men were randomised. The paper’s abstract says so, but its own flowchart shows forty-five recruited, six withdrawing before the start and thirty-nine randomised, of whom thirty-eight completed. The entry now gives the flowchart’s numbers. The trial’s finding, and its place in the second list, are unchanged.
Later the same day, reviews of our berberine and apple cider vinegar verdicts found two more entries out of date. We listed berberine’s GLP-1 effect as never measured in a person. A double-blind trial published in 2021 measured it in 300 adults with raised blood sugar over 16 weeks, and berberine moved it no more than a placebo did, so the entry has moved from the first list to the second, as this page said it would. We said nobody had tested today’s apple cider vinegar gummies; a testing company measured their acetic acid in 2020. What nobody has tested is whether a gummy does anything for weight or blood sugar, so that entry stays in the first list, reworded.
A review of our lifting verdict the same day found one more. Our entry for the arm-curl experiment described it as the same curls done once with a hormone surge and once without. Each man trained both arms, one of them followed straight away by leg work that raised hormone levels. The entry also left out two small experiments from other labs that saw a strength edge with the hormone rise. It now says both, and stays in the second list, because neither experiment showed the surge is what builds the muscle.
Our Blue Zones verdict was corrected on 18 September 2026 after an independent editorial review, and until today this page repeated a line it corrected. The entry here said one of the ten regions “has never been studied at all”. The review behind it found no study of longevity in Loma Linda as a place. The Seventh-day Adventists the town is known for have been studied, across California, and could expect to live years longer than other white Californians. The entry now says Loma Linda has never been studied as a Blue Zone, and adds that two of the three well-documented regions, Okinawa and Nicoya, are losing their advantage. It stays in the first list: most of the places that carry the name still lack the data.
On 30 September 2026 we audited every sentence on this site that says nobody has tested something, and the first list on this page failed more often than it passed. That list had 18 entries, each naming a study we said did not exist. For 12 of them the study, or part of it, was there to be found: in PubMed, where our searching had missed it; in a journal PubMed does not index; in a company’s test report; behind a review’s “none” that was out of date or wrong; or on our own verdict pages, which this page had not kept up with. The lists now stand at 12 and 8. They stood at 18 and 4.
Two entries have moved to the second list, because the test had been run and came back empty. Castor oil packs: we said we could not find a study of whether castor oil crosses skin. A 2005 study measured it in three volunteers and found no sign that it did. Cold plunges: we said nobody had measured what a plunge delivers. A 2025 experiment measured half an hour in cold water, ten times the length of a plunge: about 21 calories more burned than in half an hour of warm water, and about 230 more eaten at the next meal.
Two entries were half right and have been split between the lists. Food combining: we said its digestion promise had never been tested, “not negatively, not once”. Its mechanism was tested in 1934 and again in 1936 and failed both times, and that half is now in the second list. What we have not found is a trial of bloating, so that half stays. Lymphatic drainage: we said there were no trials at all in healthy adults for weight, contouring, detox or bloating. Weight was measured in two small studies and did not change significantly, and those are now in the second list. For contouring we found one randomised trial, in 60 women with cellulite, where fat folds were thinner after all three kinds of massage it compared and nobody went untreated [35], and one uncontrolled study that called drainage “not effective” on its own [26]: an unsettled answer, which belongs on neither list. Detox and bloating stay, and the entry now names four small trials that measured lactate or muscle enzymes after exercise, three of which reported lower levels with drainage.
One entry lost half of its claim to neither list. Oil pulling: for whitening we said there was nothing. Tooth colour has been measured in at least three published studies, and the answers do not agree. In 2021, 60 women swished coconut oil or a peroxide whitening mouthwash for two weeks: the authors report a colour change with the oil and a larger one with the mouthwash, and nobody went untreated [28]. In 2026, 100 dental students rinsed with one of three oils, a chlorhexidine mouthwash or plain water for four days without brushing: an instrument found that the oils changed tooth colour about as much as water did, and in almost everyone the teeth ended darker and yellower [6]. Also in 2026, in a trial of one maker’s oil-pulling rinse in 63 adults, teeth were on average about one step lighter on a dentist’s shade guide after 15 days and slightly darker on a placebo rinse and on an ordinary mouthwash; shade was judged by eye, it was not what the trial was built to test, and its authors call the result exploratory [7]. The maker’s own website describes a six-week study of 30 people with no comparison group [29], and an eight-week whitening trial of the same rinse, run with the maker, was registered in June 2026 and has posted no results [30]. A claim that has been measured with answers that disagree is neither of this page’s two kinds, so whitening has come off the first list and has not gone onto the second. The toxins half stays.
Four entries have left the page, because the study exists and did not come back empty either. Creatine gummies: we said nobody had bought a jar and measured what is in it, and called it the question we most wanted answered. A supplement company reported doing so in February 2024 and a supplement-rating firm in June 2025, both before this page was first published. Half or more of the brands each tested fell short of the label and the rest passed, which is a split result and not an empty one. Seed oils: we said seed oil added to an ordinary diet had never been trialled. Our seed-oils verdict has described two trials that did it since its correction of 18 September 2026, and since 30 September 2026 it describes several more, which added the oil, swapped it for starch or took it out for up to six months. Their results are mixed. Creatine and water retention: we said nobody had run the trial. Several have measured body water on creatine and found more of it; the review we quoted asks for better ones, not for a first one. Any cooking oil, against death: trials have put people on a diet built round a named oil and counted deaths since the 1960s: safflower oil in Sydney and corn oil in Minnesota, which our seed-oils verdict reports, and rapeseed and olive oil in Lyon, which our canola verdict now reports. Each changed more than the oil and their answers are disputed, which is not the same as missing.
Three more entries have been narrowed. Testosterone boosters: the review we cited examined the 16 best-selling testosterone products and the 16 best-selling erection products, 32 in all and not the 16 we printed, and its “none” is about those; a few other finished products have been trialled. Intermittent fasting: Cochrane’s search ended in December 2019, and a 2024 pilot has since counted a death and two further coronary events among 42 heart-attack patients. It is far too small to answer the question, and we found no trial that is large enough. Blue Zones: on the morning of 30 September 2026 we rewrote this entry to say that Loma Linda “has never been studied as one”, and said so in the note above this one. That sentence was ours, it was written that day, and it was wrong: several studies have examined Loma Linda as a Blue Zone, one of them comparing 1,032 hip-fracture patients from inside and outside it. What is missing is a study showing that the people who live there outlive their neighbours, and a 2025 paper by the demographers who validate Blue Zones lists the town among the places where validation did not succeed.
The other six entries held when we searched again. They now say where we looked and on what date and, where there is one, name the nearest study we found. The counts in the two headings are taken from the lists. “There is no paper to find” and the claim that the gap between the lists was “the finding” have been rewritten, because behind twelve entries there was a study, or part of one. The sea moss entry now also answers a study that a seller’s blog describes and that we could not find in the journal it names. We have also removed a figure of 74,000 monthly searches for creatine gummies, which we could not trace to a saved data pull. The studies behind the changed entries are listed under Sources.
Corrected again on 30 September 2026, after our oil-pulling verdict was corrected the same day. The oil-pulling entry said the trials behind that verdict measure plaque, gum inflammation, bacteria and breath. Between them they also measured dry mouth, saliva chemistry and tooth decay, and the entry now says so. It called drawing out toxins one of the two claims doing most of the selling; our sample of videos shows the claim being made, not that it sells more than the others, and those words have gone. And the source entry for the maker’s rinse trial gave only the paper’s declaration of no financial support; the trial’s registration names the rinse’s maker as its sponsor and source of money, and the entry now says that too. The lists are unchanged.
Corrected again on 30 September 2026, after our lymphatic drainage verdict was corrected the same day. The lymphatic drainage entry counted four small experiments after hard exercise, three of them with lower lactate or muscle enzymes after drainage. The verdict’s reading of the literature counts seven, reported in eight papers: five with lower levels than rest, two with no difference, and in the one that also tried an ordinary massage, no difference between the two. The entry now says so, and it no longer describes the reviews behind the verdict as pooling only people with lymphedema: two of the three also pooled women who did not yet have it. The weight entry in the second list now includes a third small study, from Korea. Four sources are added. The lists are unchanged.
Corrected again on 30 September 2026, after our creatine gummies verdict was corrected the same day. The creatine gummies entry said the first company test found “significant amounts” of creatinine, the substance creatine breaks down into, which read as if breakdown explained why six products fell short. The company’s own release, which we had said would not open for us, opens in the Internet Archive, and it concludes otherwise: creatine and creatinine together still fell short of the labels, so those products had apparently not been given enough creatine in the first place. The entry and its source now say so. The lists are unchanged.
Corrected again on 1 October 2026, after our castor oil verdict was corrected the same day. Source [31], the 2005 study of whether castor oil crosses skin, named the organisation that part-funded it and did not say that the organisation sells castor oil pack sessions through its spa, a stake in the answer the study was asking about. The entry now says so, and describes the funder by its role, as our other pages do. The study found no sign that the oil gets through the skin, so its finding ran against that stake. The lists are unchanged.
Corrected again on 1 October 2026, after our walking-after-eating verdict was corrected the same day. The first list said we had looked for trials of walking after meals and body weight and found none, and named as the nearest things a 2011 report of two people [13] and a trial in pregnant women withdrawn before anyone enrolled [14]. The trials exist. In a 2017 Norwegian trial, people who added at least 30 minutes of light activity starting within half an hour of a meal, every day for 12 weeks, lost a little weight while the comparison group gained a little, a difference its authors report as significant; and a 2016 crossover trial weighed and measured people on two walking plans and found no difference in body mass index or waist over two weeks. The test exists and did not come back empty, so, as this page says it will, the entry has left the page: the first list now has 11 entries, and its footer has a different example. The second list is unchanged.
Corrected on 7 October 2026, after our beef tallow verdict was corrected the same day. The tallow entry quoted only the second half of a 2024 scoping review’s conclusion, the call for research on tallow as a cosmetic product in people; its first half found tallow creams moisturising in small tests, and the entry now gives both. It also described the nearest studies as two, with walnut oil; the review holds three small tests from one research group, and two of the creams used other plant oils. The entry stays in the first list: none of the three set a tallow cream against one you could buy, and we still found no trial of tallow as it is sold.