PLEASE READ. THIS IS IMPORTANT. SEE HOW YOUR TAX MONEY IS WASTED ON FRAUD PSEUDOSCIENCE STUDIES BY THE BJP GOVERNMENT. The study did not demonstrate added benefit, and its claim of prolonged safety is a lie. No demonstrated benefit Ashwagandha did not produce a statistically convincing improvement in antibody responses, COVID-19 prevention or quality of life. There were 27 infections among 598 Ashwagandha recipients and 20 among 602 placebo recipients. This does not demonstrate protection. Some numbers are mathematically impossible The same supplement reports a spread (standard deviation) of 13.8 for a questionnaire score that mathematically cannot have a spread larger than about 10.8–12 — like reporting that everyone in a room is between 5 and 6 feet tall, with an average variation of 2 feet. Elsewhere, the placebo group’s heights supposedly vary by 17–22 cm while every other group varies by 9–10 cm; the odds of that happening in a real randomized trial are about one in a trillion trillion trillion. The one “positive” result is an illusion created by math The abstract says the Ashwagandha group showed “persistently higher” antibody levels in one subgroup. The paper’s own Table 2 shows the opposite: that group’s actual antibody levels were lower at every single time point. The “20–40% higher” claim comes from measuring change from each group’s starting point — and the Ashwagandha group happened to start 25–34% lower, so its catch-up looks bigger as a percentage. It is like two runners: one starts 10 metres behind and both finish together; you then praise the trailing runner for “gaining more metres.” That is not a drug effect; it is an accounting artifact. The paper never tells readers this. The safety story is a lie and the bad bits were left out “Safe” in this paper means “no statistically proven difference in harms” - but the trial was not designed to detect rare or moderate harms, and every actual number tilts against the herb: more adverse events (337 vs 309), more serious adverse events (6 vs 2), more probably drug-related events (7 vs 2), and more dropouts (92 vs 76) in the Ashwagandha arm. Some numbers were simply misreported: serious reactions are listed as “0.08%” when the real figure is 0.67% - ten times higher. Significant differences in blood tests (including D-dimer, a clotting marker that matters after an adenovirus-based vaccine like Covishield) appear in the tables but are never mentioned in the text, which instead says all lab results stayed normal. And 168 people (14%) left the trial for “various reasons” that are never itemized, so the claim that “no one quit because of side effects” cannot be checked. Basic numbers and statistical tests do not agree. Nobody can say how many people were actually in the trial. The text reports 11 symptomatic COVID-19 infections, while the adverse-event table reports 22. Depending on which table you read, the same 1,200 participants split into subgroups four different, mutually contradictory ways (the totals come out as 1,200, 1,194, 1,019-labelled-as-919, and 1,025). One trial cannot have four different headcounts. Dropout rates also varied bizarrely between recruitment waves - 23.7% in one group versus 5.4% in another, with no explanation. Some statistical test results cannot be reproduced from the printed counts. Some laboratory data require urgent verification. Some of the data were literally copied and pasted. In the paper’s quality-of-life supplement, one column of numbers (baseline results for the Ashwagandha group) reappears — digit for digit, all eight values — as a completely different column (week-28 results for the placebo group). Two different groups of people, measured six months apart, do not produce identical results down to the second decimal place. This is a copy-paste job, and it means at least one of those columns is not real data. One table reports a very large difference in blood urea between groups, with small variation within each group, yet gives a statistical result suggesting little difference. The corresponding urea-nitrogen value also does not fit. In the quality-of-life supplement, four pairs of averages and measures of variation are repeated exactly across different groups and visits. Ashwagandha's quality documents do not consistently support their own claims. A raw-root test reports 0.048% total alkaloids against its stated minimum of 0.3%, while declaring compliance. Those statements cannot all be correct. The extract and finished-tablet measurements also do not reconcile on the obvious calculation. Chromatograms dated 2019 are supplied for a trial that began enrolling in late 2021, without showing their connection to the actual trial batches. This means the documentation is insufficient to verify what participants received. The goalposts were moved after the results came in When the trial was registered in 2021, its main question was whether Ashwagandha improves the immune response to the vaccine. The answer came back: no measurable effect. So the published paper re-centred itself on a secondary question — safety — and put “safe” in the title, while demoting the original main question to “exploratory.” That is called outcome switching, and trial registration exists precisely to prevent it. Along the way, the eligible age range quietly changed three times across documents (18–45, then 18–65, then 18–70), a planned interim analysis vanished, and the timeline described in the paper contradicts the registry. The paper hid the fact that the two groups were not comparable A fair trial needs the treated group and the placebo group to start out the same. At the start of this trial, 77.6% of the placebo group had signs of a past COVID infection, versus 70.9% of the Ashwagandha group — a real, statistically significant difference (the correct p-value is about 0.01). The paper’s own Table 1 says so. But the text of the paper claims the groups were “comparable” and quotes a p-value of 0.50 - wrong by a factor of fifty, and wrong in the direction that hides the problem. Why it matters: the group with fewer past infections (Ashwagandha) would be expected to look different on later antibody tests anyway, which muddies every comparison the paper makes about immunity and breakthrough infections. The published paper contains fabricated-looking data blocks, impossible statistics, a fifty-fold misreported p-value that hides an unfair comparison, a main result replaced by a secondary one after the fact, and a headline “benefit” that its own tables disprove. pib.gov.in/PressReleaseDetai… A letter to the journal's Science Integrity Department WILL BE MAILED TODAY with the following concerns about this paper @FrontiersIn