✶ Written by an AI · fact-checked before publication
Every claim in Adaptogens, Checked, checked
This is the book's verification ledger, published in full. Every factual claim was extracted from the manuscript and handed to an independent blind checker — one claim, its cited source, nothing else. High-stakes claims (doses, safety, real diseases, trial numbers) were checked by three checkers who had to agree. Every checker first had to confirm the cited source actually exists.
72claims checked
59verified first pass
13flagged & corrected
The honest part: 13 claims did not survive first contact with their own sources — the checkers flagged them as misstated or unsupported. Each one was corrected in the text and re-verified by a fresh blind panel before the book was published. The table below shows the first-pass verdicts, because that's the honest number: what the machine got wrong before checking, not after.
| id | claim as written | stakes | first-pass verdict |
|---|---|---|---|
| C1 | James Whitfield is a large language model — a piece of software that predicts words; not a ghostwriter, pen name, or human. | low | unsupported → corrected |
| C2 | Every factual claim in the book was checked against its primary source; where the machine reached past the evidence or invented a citation the check caught it and it was cut/corrected; a practising doctor reviewed the results and is accountable. | low | verified |
| C3 | A national medicines regulator put a liver warning on the single most popular herb in the category — ashwagandha. | high | misstated → corrected |
| C4 | The book is general education only — not medical advice, diagnosis, or treatment; do not start/stop/change any medication because of it. | low | verified |
| C5 | In 2025, researchers pulled 558 titles from Amazon's "Herbal Remedies" category and ran AI-detection analysis; roughly 82% came back as likely AI-written. | high | verified |
| C6 | The single top-ranked book in that category was credited to an author who does not appear to exist — "Luna Filby," with no findable human behind it. | high | verified |
| C7 | Around 76% of the fake-seeming authors used names drawn from the same small set (Rose, Fern, Sage — botanical-cottagecore). | high | misstated → corrected |
| C8 | Roughly 54% of the blurbs were sprinkled with leaf emojis. | high | unsupported → corrected |
| C9 | One such book claimed echinacea had "resolved" chronic infections, presented as settled fact with no dose, duration, contraindication, or interaction. | high | verified |
| N1 | The doctor is a practising, registered Australian clinician who stays unnamed on purpose. | low | unsupported → corrected |
| N2 | Corrections can be reported at jameswhitfieldauthor.com/corrections and will be fixed in the open. | low | verified |
| N3 | Adaptogens more broadly can stack with sedatives, thyroid medication, immunosuppressants, blood-sugar-lowering drugs, and blood thinners. | high | verified |
| N4 | Millions of doses of ashwagandha are taken, so 12 reports against that backdrop is statistically rare. | high | unsupported → corrected |
| C10 | Other books contained recipes calling for ingredients missing from their own ingredient lists and steps that trailed off mid-instruction. | low | verified |
| C11 | Some cited publicly discredited figures — Barbara O'Neill and "Dr. Sebi" — as though they were reliable sources. | high | verified |
| C12 | The book itself is machine-written (openly); the real fault line in the category is checked-vs-unchecked and honest-about-risk vs quiet-about-risk, not human-vs-machine. | low | unsupported → corrected |
| C13 | In the 1940s a Soviet pharmacologist, Nikolai Lazarev, coined "adaptogen" for a substance that raises the body's "non-specific resistance to stress"; his work was carried forward by Israel Brekhman, whose name is on most foundational adaptogen research through the 1960s. | low | verified |
| C14 | Brekhman and his colleague Dardymov defined "adaptogen" by three criteria: non-specific (broad resistance), normalising (two-way corrector toward balance), and non-toxic/non-disruptive at normal doses. | low | verified |
| C15 | Adaptogen" is not a regulated or clinically validated drug class; no medicines regulator anywhere (not TGA, FDA, or EMA) licenses a drug as "an adaptogen. | high | verified |
| C16 | Many herbs sold as adaptogens have no evidence they meet even the historical criteria; independent reviewers keep finding the marketing overstates the evidence and for many claimed properties there is no good support. | high | verified |
| C17 | A cortisol number moving in a trial is not the same as a person's life feeling less stressful — cortisol is a surrogate marker, the real outcome is harder to measure and the two don't move in lockstep. | high | verified |
| C18 | Much of the adaptogen literature consists of small, short, frequently industry-funded studies leaning on self-reported stress scores — the combination that makes evidence weak. | high | verified |
| C19 | Not shown" means absence of good proof, not proof of absence — not the same as "shown not to work. | low | verified |
| C20 | Language models are fluent and confident even when wrong and can generate a beautifully worded, completely made-up citation — the known failure mode. | low | verified |
| C21 | Ashwagandha is sold for stress, sleep, cortisol reduction, "testosterone"/hormonal support, and strength or muscle. | low | verified |
| C22 | Several small RCTs report people taking ashwagandha scored lower on self-reported stress scales and showed reductions in measured cortisol vs placebo; the trials are mostly small (often a few dozen), frequently short (a couple of months), and a striking number are funded by or linked to the companies selling the product. | high | verified |
| C23 | Independent appraisers found no robust, randomised, double-blind support for the metabolic/hormonal claims (blood sugar, energy, testosterone). | high | misstated → corrected |
| C24 | On 22 February 2024, Australia's medicines regulator, the TGA, issued a safety alert on ashwagandha over liver toxicity. | high | verified |
| C25 | That alert followed 12 reports of liver problems and 4 hospitalisations, and Australian products containing ashwagandha are now required to carry a liver-injury warning statement. | high | misstated → corrected |
| C26 | Verdict: modest short-term self-reported stress-and-cortisol signal is the most defensible claim; hormonal/metabolic promises not robustly supported; safety-flagged overall. | high | verified |
| C27 | Rhodiola is sold for fatigue, mental performance under pressure, and general stress resistance. | low | verified |
| C28 | A handful of short studies found improvements in fatigue and perceived stress, but the trials are generally small, heterogeneous (different doses, preparations, outcome measures), and of mixed methodological quality; the real-life effect stays uncertain. | high | verified |
| C29 | Verdict: mixed-to-weak; across the five it comes out one of the better-tolerated, but the case it reliably lifts fatigue or stress isn't proven. | high | verified |
| C30 | Panax ginseng has been used for a very long time and is sold for energy, cognition, and "vitality. | low | verified |
| C31 | Ginseng has a large evidence base of widely varying quality; weighed together it shows small and inconsistent effects on fatigue and cognition, and product standardisation (ginsenoside content, preparation) is a major confounder. | high | verified |
| C32 | Ginseng carries real interaction cautions: with anticoagulant/antiplatelet drugs (added bleeding risk) and with antidiabetic drugs (added blood-sugar lowering). | high | verified |
| C33 | Tulsi/holy basil is sold for stress and vague "metabolic" benefits. | low | verified |
| C34 | A handful of small trials report self-reported stress improvements at low certainty. | high | verified |
| C35 | Tulsi shows a blood-sugar-lowering signal in the research; this is a safety/interaction consideration (caution with diabetes medication, additive lowering), not a green light to treat it as a blood-sugar remedy. | high | verified |
| C36 | Eleuthero — sometimes sold as "Siberian ginseng," though not a true ginseng — was the original herb of the Soviet adaptogen research. | low | misstated → corrected |
| C37 | The evidence for eleuthero on endurance and fatigue is thin, dated, and largely low-quality. | high | verified |
| C38 | Eleuthero has historically been frequently adulterated or mislabelled; product analyses turned up supplements that weren't reliably eleuthero. | high | verified |
| C39 | Even where a trial found a positive effect, the effect tends to be modest and short-term (a few weeks, not years). | high | verified |
| C40 | The testosterone/hormonal claim is where marketing most outruns the data — the gap is widest exactly where the label is loudest. | high | verified |
| C41 | The adaptogen literature is dominated by small, short-window trials, often funded by parties with a stake in the answer, measuring soft self-reported endpoints; independent replication is scarce. | high | verified |
| C42 | Managing load and recovery has reasonable evidence and is partly known (harder to trial cleanly); this list is general wellbeing, not treatment for any diagnosed condition. | high | verified |
| C43 | On 22 February 2024 the TGA (Australia's national medicines regulator) issued a safety alert on ashwagandha (*Withania somnifera*) over hepatotoxicity — a live regulator action on the public record. | high | verified |
| C44 | The alert followed 12 reports of liver problems and 4 hospitalisations. | high | verified |
| C45 | Australian listed products containing ashwagandha are now required to carry a liver-injury warning statement. | high | misstated → corrected |
| C46 | St John's Wort induces drug-metabolising enzymes, reducing the effectiveness of many prescription drugs. | high | verified |
| C47 | Concretely, it can make some hormonal contraceptives less reliable, blunt anticoagulants, and undercut immunosuppressants. | high | verified |
| C48 | Combined with certain antidepressants, St John's Wort adds serotonin-syndrome risk. | high | verified |
| C49 | Adaptogens with a calming/sedating signal can add to sedative medications (additive drowsiness). | high | verified |
| C50 | Ashwagandha has a thyroid signal and can interact with thyroid medication. | high | verified |
| C51 | Adaptogens can interact with immunosuppressant drugs. | high | verified |
| C52 | Tulsi and ginseng both carry a blood-sugar-lowering signal that can add to antidiabetic medication. | high | verified |
| C53 | Ginseng carries an anticoagulant/antiplatelet interaction — a bleeding-risk flag on top of a blood thinner. | high | verified |
| C54 | Ashwagandha in particular is traditionally cautioned against in pregnancy. | high | verified |
| C55 | Safety in pregnancy and breastfeeding is not established for this set of herbs; the honest default is avoid-and-ask. | high | verified |
| C56 | In Australia, complementary medicines mostly sell as *listed* medicines, which are not assessed by the TGA for efficacy before sale; the TGA checks permitted ingredients and manufacturing quality, not whether the product works, and the "AUST L" number reflects paperwork filed, not efficacy. | high | verified |
| C57 | Dose differs between products, the actual plant species can differ, and adulteration/mislabelling are documented — eleuthero (Siberian ginseng) has a long history of substitution or mislabelling. | high | verified |
| C58 | Talk to a GP or pharmacist BEFORE combining any of these herbs with a prescription medication, using them in pregnancy/breastfeeding, or taking them with liver disease. | high | verified |
| C66 | The McGill Office for Science and Society, with no supplement to sell, stated the studies needed to back the glucose, energy, and testosterone promises largely don't exist. | high | verified |
| C67 | The six promises the adaptogen shelf makes are: reduces cortisol/lowers stress; boosts energy/stamina; raises testosterone/hormonal support; improves sleep; boosts immunity; "adapts" the body (dials whatever's out of whack back to normal). | low | verified |
| C68 | Of those six, only a handful survive appraisal and only in a modest short-term self-reported form; the testosterone/hormonal, energy/stamina, immunity, and "adapts the body" claims are where marketing most outruns the data. | high | verified |
| C69 | Scoreboard verdicts: cortisol/stress = mixed-to-weak; energy/stamina = not shown; testosterone/hormonal = not shown (no robust double-blind support); sleep = mixed-to-weak; immunity = not shown; "adapts the body" = not a testable claim. Of six promises, none is strongly supported. | high | misstated → corrected |
| C70 | Regular movement/physical activity has strong evidence for stress, mood, and stamina — one of the best-evidenced things for the cluster of complaints adaptogens are marketed at. | high | verified |
| C71 | Consistent sleep-wake timing has strong evidence for sleep quality and daytime energy; morning daylight has good evidence for sleep timing and daytime alertness. | high | verified |
| C72 | For the outcomes adaptogens are marketed for, the adaptogens themselves are at best a modest short-term nudge on self-reported stress for one or two herbs; everything else is not shown, and where the honest answer is "we don't fully know" that is the answer. | high | verified |
| C73 | Myth #1: "Natural" does not mean "safe" — a herbal product can still cause serious harm, and being on a shop shelf does not make it harmless. | high | verified |
| C74 | The five myths are ranked worst-first by how badly a reader can be hurt: liver injury and missed drug interactions at the top, label-and-efficacy gaps at the bottom. | high | verified |
| C75 | The greatest harm in the category flows from the "natural = safe" belief because it is the gateway to ignoring every other caution. | high | unsupported → corrected |
Ledger exported 2026-07-23. Found something we still got wrong? Tell us — corrections happen in the open. How the books are made: the honest method.