# COVID vaccine narrative map for Getting It Right

Research cut-off: 14 October 2026. Research and editorial recommendations only; nothing published.

## Decision summary

1. **Best new topic: COVID vaccination, cancer and DNA.** The gap is real even with the site's existing COVID, deaths, harms, guidance and myocarditis pages. Cancer/DNA claims recur across several years, languages and platforms, appear in political and advisory settings, and were documented in Indonesia as recently as 7 October 2026. Put cancer and DNA in one coherent page initially; cross-link rather than create separate pages for every proposed mechanism.
2. **Next distinct topic: COVID vaccination, fertility and pregnancy.** There is older representative evidence of belief, plus concrete 2026 recirculation of miscarriage and infant-death claims. This merits more than a passing mention, with fertility, miscarriage, menstrual changes, birth rates and infant outcomes distinguished. Coordinate with any broader vaccine/pregnancy page to avoid duplication.
3. **Update existing deaths, myocarditis and harms pages first for scientific precision.** Real rare harms are not misinformation. The 2025–26 pediatric death review, persistent cardiac MRI abnormalities, some product-specific IVF findings and new IgG4/infection associations must not be erased to make a reassuring story simpler.
4. **Immune-collapse claims deserve a substantial section, plus research watch.** IgG4 switching and exploratory post-vaccination illness research are real. The jump to generalized immune destruction, AIDS or proven cancer causation is unsupported. Current prevalence is not quantified well enough here to justify another standalone page immediately.
5. **Keep outbreak-specific variants on watch.** Hantavirus, tuberculosis, food-borne vaccine and secret-shedding variants should feed explanatory sections, not a growing list of thin pages.

These are editorial priorities based on recency, consequence, evidence and existing coverage. They are **not a measured worldwide ranking of misinformation prevalence**.

## Scope and method

This was a structured, multi-pass rapid evidence map, not a registered systematic review. Searches covered: (1) representative surveys; (2) dated original statements, documents and testimony; (3) recent fact-check archives; (4) multilingual recurrence; (5) population safety and effectiveness studies; (6) laboratory/mechanistic evidence; (7) the strongest contrary papers cited by claimants; and (8) correction/retraction notices and follow-up outcomes. Search-result dates were checked against publication dates inside sources where available.

Publicly indexed sources were available. Private messaging groups, comprehensive platform firehoses, deleted posts and reliable cross-platform unique audience counts were not. English-language US/UK sources dominate, with verified examples in French, Indonesian, Malay and German; this is not representative global surveillance. A source appearing in another language establishes cross-language recurrence, not national prevalence. Shared appearances, citations and repeated talking points do not by themselves establish a covert coordinated campaign or intent to deceive.

The parent verified these existing routes: /topics/vaccines/, /topics/covid/, /topics/mmr/, /topics/vaccines-autism/, /topics/covid-myocarditis/, /topics/covid-vaccine-deaths/, /topics/covid-vaccine-harms/, /topics/covid-vaccine-guidance/, /topics/vaccine-amplification/. Recommendations below respect that inventory; this report did not audit their exact prose.

### How prominence is measured here

- **Exposure:** prompted self-report of having heard/read a claim; not agreement and not necessarily recent exposure.
- **Belief:** exact survey response category, especially the distinction between definitely and probably true.
- **Reach:** a recorded view/share/access count with date and platform; these are not unique people or belief.
- **Recurrence:** independently documented examples at different dates/places; cannot be converted into population frequency.
- **Institutional amplification:** an official hearing, guidance document or advisory presentation. This establishes a consequential venue, not audience size or scientific validity.
- **Harm:** plausible decision consequences are editorial assessments unless a study directly measures them. No causal estimate of deaths attributable to these individual narratives is made.

## Prominence anchors

- **Current US survey:** KFF's 30 June 2026 release, fielded 7–31 May among 2,480 adults, reports 46% had heard that COVID vaccines killed more people than the virus and 36% had heard that mRNA vaccines alter DNA. Fewer than 10% called each of the four tested myths definitely true. DNA exposure fell from 45% in April 2025; do not claim every narrative is rising. Among parents, 35% said the deaths claim was probably/definitely true, and 29% said this about DNA. This is prompted US opinion, not worldwide online volume. [KFF release and methodology](https://www.kff.org/public-opinion/kff-tracking-poll-on-health-information-and-trust-update-on-common-vaccine-myths/)
- **Historical US baseline:** KFF's May–June 2023 survey found 65% had heard the sudden-deaths claim. The topline records 34% probably/definitely true for thousands of sudden deaths, 27% for proven infertility (only 3% definitely true), and 20% for more deaths from vaccines than virus. Wording and dates matter; do not substitute these figures for 2026 belief. [Report](https://www.kff.org/health-information-trust/kff-health-misinformation-tracking-poll-pilot/) · [Original topline](https://files.kff.org/attachment/Topline-KFF-Health-Misinformation-Tracking-Poll-Pilot-August-2023.pdf)
- **General confidence is different:** Annenberg's November–December 2025 US panel of 1,637 adults found 65% regarded COVID vaccines as safe, compared with 73% in August 2022; no statistically significant change from November 2024. This measures safety perception, not assent to a particular narrative. [Annenberg, 10 February 2026](https://www.annenbergpublicpolicycenter.org/study-finds-declining-perception-of-safety-of-covid-19-flu-and-mmr-vaccines/)

## Editorial matrix

| Narrative family and variants | Geography/language actually evidenced | Prominence and recency | Main risk to readers; evidence judgment | Recommended home |
|---|---|---|---|---|
| Vaccines caused mass deaths; VAERS equals confirmed fatalities; excess mortality proves depopulation; Japan declared the deadliest drug | US/UK English, Canada-origin report; Spanish translation documented | Current KFF exposure; 2024 Carlson/Weinstein amplification; Japan-authority claim shared thousands of times again in August 2026 | Refusal of useful protection; bereaved families' losses misattributed. Strong evidence against claimed mass mortality; rare fatal harms must be acknowledged | Update existing deaths and harms pages; explanatory VAERS section, no duplicate |
| All young hearts are permanently damaged; athlete deaths prove it; myocarditis means vaccination is always worse | US/UK/Nordic English scientific and public discussion | Real FDA warning June 2025; death-review controversy 2025–26. No current claim-specific prevalence estimate established here | Both inflated risk and minimizing a real adverse effect harm trust and care. Causal mRNA–myocarditis relationship strong; individual net benefit conditional | Existing myocarditis page, with honest outcome and subgroup boxes |
| Vaccination causes aggressive/turbo cancer, recurrence or a cancer epidemic | English US/UK/Canada; French; Indonesian; imported Korean/Italian papers | September 2025 political conference and ACIP presentation; June 2026 Senate hearing; September/October 2026 social recurrence | Cancer patients may avoid protection or confuse treatment decisions. No established causal cancer link; positive observational findings require direct treatment | Highest-priority new combined cancer/DNA page |
| Residual DNA, SV40 or reverse transcription rewrites the genome; heritable damage | US official English guidance, international English; Indonesian crossover | KFF 2026 prompted exposure 36%; Florida January/September 2024; ACIP September 2025; current cancer recirculation | Misunderstanding manufacture becomes permanent-mutation fear. Residual DNA is real, harmful genomic integration in recipients not demonstrated | Major section of cancer/DNA page; short mRNA explainer linked from umbrella |
| Infertility, miscarriages, birth collapse, disabled babies or infant deaths | US/UK English; French 2026 recirculation; Czech/Swedish/Chinese studies cited internationally | Older KFF belief; April 2026 infant-death post; August 2026 Fauci-text/82% miscarriage revival | High-stakes pregnancy decisions. Substantial reassuring mRNA evidence; menstrual changes real; some inactivated-vaccine ART signals unresolved | Second new topic candidate; connect with guidance, avoid universal pregnancy policy claims |
| Immune destruction, VAIDS, IgG4 cancer tolerance, perpetual spike poisoning | English; German August 2026; Malay February 2026 | ACIP September 2025; 2026 research recirculation; no representative belief estimate found | Confusing immunological measurements with generalized illness; invalidating genuine symptoms also harms patients. Specific IgG4 associations real; AIDS/cancer leap unsupported | Substantial harms/cancer section; watch emerging science |
| Vaccines do not work because infection/transmission still occur; more doses prove susceptibility; natural immunity makes any dose useless | US official English policy statements; international online versions | HHS August 2025 announcement justified termination/de-scoping of 22 projects; ongoing scientific disputes | Treating infection as the only outcome hides serious-disease benefit. Effectiveness varies by outcome, season, prior immunity and risk | Update existing COVID and guidance pages, benefit-first |
| Outbreak-specific contaminants: TB or hantavirus; self-amplifying RNA spreads uncontrollably | Malay/English; other versions require separate checks | Malaysia February 2026; hantavirus May 2026; specific current falsehoods, no population reach estimates | Distracts from actual infection prevention; spills into other vaccination | Watch/examples within ingredients/immune sections; no new standalone page yet |

## 1. Deaths, VAERS and excess mortality

### Claim provenance and observable technique

On 5 January 2024 Bret Weinstein told Tucker Carlson that a roughly 17-million vaccine-death estimate was credible. The cited September 2023 Correlation report extrapolated from mortality in selected Southern Hemisphere countries. The methodological issue is attributing post-rollout excess mortality to vaccination without adequately separating infection and other causes, then extrapolating globally. This is a causal and denominator problem, not a reason to disregard excess-mortality data altogether. [Dated reporting](https://www.factcheck.org/2024/01/scicheck-tucker-carlson-video-spreads-falsehoods-on-covid-19-vaccines-who-accord/) · [Archived original interview](https://web.archive.org/web/20240108035213/https://tuckercarlson.com/the-tucker-carlson-encounter-bret-weinstein/) · [Archived original report](https://perma.cc/MLU9-XTQL)

A current variant borrows national authority: Full Fact on 27 August 2026 documented posts shared thousands of times claiming Japan declared mRNA COVID vaccines the deadliest drug. It traces them to a November 2024 People's Voice article and a politician's conference, not a Japanese government determination. This is evidence of repeated sharing, not a count of believers. [Full Fact](https://fullfact.org/world/japan-covid-deadly-vaccine-false-claim/)

### Correction and strongest caveats

Suggested plain lead: **COVID vaccines reduce the risk of severe COVID. Large studies do not support claims that they caused mass deaths. Rare fatal reactions can occur, and every suspected death deserves investigation.**

- A French national cohort compared 22.7 million mRNA-vaccinated and 5.9 million unvaccinated adults aged 18–59, with median follow-up of 45 months, and found no elevated all-cause mortality. It also observed lower severe-COVID mortality. Observational selection and healthy-vaccinee confounding mean lower all-cause mortality should not be advertised as proof vaccination prevents every cause of death. This age range and study period do not answer all pediatric, product or future questions. [JAMA study](https://doi.org/10.1001/jamanetworkopen.2025.46822) · [French study team's accessible summary](https://www.assurance-maladie.ameli.fr/etudes-et-donnees/2025-vaccin-covid-arn-messager-mortalite-quatre-ans)
- An English self-controlled study of ages 12–29 found no significant overall increase in cardiac/all-cause death after vaccination, but did find a cardiac-death signal in women after a first non-mRNA dose. Authors discuss clinical vulnerability and selection; the result cannot be generalized to all current mRNA products. [Nafilyan et al., Nature Communications 2023](https://www.nature.com/articles/s41467-023-36494-0)
- A Korean adjudicated surveillance study identified 480 vaccine-related myocarditis cases among 44.3 million vaccinated people, including 21 deaths. Ascertainment and case classification limit generalization, but these findings make “no vaccine-related deaths” an indefensible reassurance. [European Heart Journal 2023](https://academic.oup.com/eurheartj/article/44/24/2234/7188747)
- **Important 2026 update:** the December 2025 FDA memorandum, publicly released in May 2026, reviewed 96 US pediatric death reports. It classified two as probably related, five possibly related, 62 unlikely and 27 unassessable; none certain. All seven possible/probable cases followed Pfizer formulations, ages 7–16, and involved cardiac conditions; five involved myocarditis. “Possible” permits an equally likely alternative cause. Zero “certain” is not proof zero vaccine-caused deaths; seven possible/probable is not seven proven deaths, and the dataset cannot establish an incidence rate. The memorandum explicitly warns about these interpretive problems. [Original released memorandum, PDF pp. 8–15](https://www.ronjohnson.senate.gov/wp-content/uploads/2026/05/2026-05-11-RHJ-to-HHS-re-C19-vaccine-records1.pdf)

VAERS is a signal-detection system. A report can be important and sincere while causality remains unresolved. Underreporting does not authorize multiplying every report by an arbitrary constant; stimulated reporting, reporting requirements, duplicates, baseline event rates and missing comparison groups also matter. Explain how a signal led to confirmation of myocarditis and product-specific clotting risks, rather than using a dismissive “anyone can submit anything” slogan. The existing deaths/harms pages are the right location.

## 2. Myocarditis: retain the inconvenient facts

Suggested lead: **COVID vaccination can prevent serious illness, but mRNA vaccines can rarely cause myocarditis. The risk is highest in particular young-male groups and depends on the product, dose and timing.**

- Nordic linked registries found the highest risk after second doses among males aged 16–24, with a higher risk following original Moderna than Pfizer formulations. This is an observed product/age/sex/dose pattern, not a uniform risk for everybody. [Nordic cohort, JAMA Cardiology 2022](https://pmc.ncbi.nlm.nih.gov/articles/PMC9021987/)
- The English self-controlled analysis estimated **97 excess myocarditis events per million** men under 40 after a second mRNA-1273 dose, versus **16 per million** after a positive SARS-CoV-2 test before vaccination, within its 28-day windows. Thus “infection always causes more myocarditis than vaccination” is false. These historical estimates depend on detected infections, formulation, era and endpoint; they are not today's all-age booster risk or a complete benefit–harm calculation. [Patone et al., Circulation 2022](https://pubmed.ncbi.nlm.nih.gov/35993236/)
- FDA's June 2025 warning described **unadjusted myocarditis/pericarditis incidence** after 2023–24 formulations: approximately 8 per million doses for ages 6 months–64, and 27 per million for males 12–24, days 1–7. These are not excess-risk estimates and should not be placed beside the English 28-day estimates without explanation. FDA also noted persistent MRI injury markers in some patients, with uncertain prognostic significance. [FDA safety communication](https://www.fda.gov/vaccines-blood-biologics/safety-availability-biologics/fda-approves-required-updated-warning-labeling-mrna-covid-19-vaccines-regarding-myocarditis-and)
- A French study followed 4,635 hospitalized myocarditis patients; the 558 post-vaccination patients had fewer severe cardiovascular events than other myocarditis groups over 18 months, but some needed ongoing care. Better average outcomes are not “harmless,” and a favorable average does not predict every patient's recovery. [JAMA 2024 primary study](https://pmc.ncbi.nlm.nih.gov/articles/PMC11348078/)

Do not treat genuine myocarditis reports or a family's request for investigation as misinformation. The misleading content is extrapolation to population-wide hidden heart damage, causal attribution of unrelated athlete collapses, or a benefit–risk conclusion that ignores baseline severe-COVID risk, prior immunity and relevant vaccine options. Conversely, dismissing risk because infection can also cause myocarditis is an evasion. Keep chest pain/breathlessness/palpitations advice clinically reviewed and direct readers with symptoms to medical assessment.

## 3. Cancer and DNA: the clearest new editorial gap

### Recurrence, amplification and geographic scope

- In September 2025 Aseem Malhotra's Reform UK conference speech linked royal-family cancers to vaccination. The venue and subsequent coverage demonstrate public amplification; this report did not verify a total audience. [Full Fact, 8 September 2025](https://fullfact.org/health/aseem-malhotra-reform-conference-vaccines/)
- The September 2025 ACIP workgroup presentation raised DNA impurities, cancer, IgG4 and frameshifting. These can be checked in its original slides. Treat claims from a government-hosted presentation as claims to evaluate, not automatically as consensus or established CDC findings. [Original slides](https://www.cdc.gov/acip/downloads/slides-2025-09-18-19/06-el-deiry-kuperwasser-covid-508.pdf) · [Contextual reporting](https://www.factcheck.org/2025/11/cdc-vaccine-panel-presentation-distorts-research-on-safety-of-mrna-covid-19-vaccines/)
- A Senate hearing on 3 June 2026 was explicitly framed around mechanisms of COVID injections causing cancer. Seven listed witnesses included claim proponents and an oncology representative presenting contrary evidence. It demonstrates an institutional agenda and dissemination routes, not unanimous findings or measured reach. [Original hearing advisory](https://www.ronjohnson.senate.gov/2026/06/04/media-advisory-chairman-johnson-to-hold-hearing-plausible-mechanisms-of-covid-19-injections-causing-cancer-and-attacks-on-scientific-publications-and-researchmedia-advisory/) · [ASCO's original testimony](https://www.hsgac.senate.gov/wp-content/uploads/Gralow-Testimony.pdf)
- Full Fact documented another celebrity-cancer attribution on 21 September 2026. [Article](https://fullfact.org/health/prostate-cancer-covid-vaccine/)
- AFP on 7 October 2026 traced a 14 September Indonesian Facebook post to William Makis's claims, alongside a Nicolas Hulscher video about supposed Pfizer DNA in a heart tumor, and examples on Instagram, TikTok and X. No numeric audience measurement was supplied. [AFP's dated examples](https://factcheck.afp.com/doc.afp.com.D28N9G3)
- Historical cross-language spread is measurable in limited units: AFP's January 2023 French report recorded a Children's Health Defense broadcast shared over 1,000 times on Facebook and a particular Twitter post above 110,000 views. These are historical platform counters, not today's prevalence or unique people. [AFP French investigation](https://factuel.afp.com/doc.afp.com.336Z237)

Observable techniques include attaching a familiar celebrity's diagnosis to a speculative cause; calling an undefined cluster “turbo cancer”; treating a plausible laboratory mechanism as demonstrated clinical harm; emphasizing a statistically significant result while omitting sensitivity analyses; and using the presence of a DNA sequence as if it established infectious virus or genome integration. These descriptions concern the content, not the speaker's state of mind.

### What the page can truthfully say

Suggested lead: **COVID vaccination helps protect people, including people with cancer, from severe COVID. Current evidence has not established that it causes cancer or a new disease called “turbo cancer.”**

“Turbo cancer” is not a standard diagnostic category. Aggressive cancers existed before COVID vaccination. A newly discovered tumor after a dose is a reason to assess the patient, not sufficient evidence that the dose caused it. **Do not say there are no papers reporting an association. There are.** Separate incidence, recurrence, progression, cancer hospitalization and cancer death, which are different outcomes.

#### Strongest contrary epidemiology, addressed directly

1. **Korean 2025 correspondence:** the study started with records for 8.4 million people, with approximately 3 million in its main matched comparison, and reported higher one-year incidence of several cancers. Its publisher still displayed an October 2025 notice about concerns under investigation when checked. Do not call it retracted. The publisher displayed about 467,000 accesses and an Altmetric score of 8,895 on access; these include interested/critical attention and are not unique exposed people or believers. [Original paper and notice](https://link.springer.com/article/10.1186/s40364-025-00831-w)
2. **Why it cannot establish vaccine causation:** an independent published methodological critique identifies different index dates for vaccinated and unvaccinated groups, unmeasured screening/health-care-use differences, residual confounding and multiple comparisons. One-year follow-up is particularly poor for attributing newly initiated solid tumors. That timing criticism alone does not disprove an acceleration hypothesis; it means the proposed mechanisms need appropriate tests rather than assertion. [Locquet et al., 2026 methodological commentary](https://public-pages-files-2025.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2026.1772406/pdf)
3. **Italian 2025 cohort:** among 296,015 residents, vaccination was associated with a higher cancer-hospitalization hazard in one analysis (HR 1.23, 95% CI 1.11–1.37), but the direction reversed when a 12-month minimum lag was imposed. The authors call findings preliminary because healthy-vaccinee bias and unmeasured confounding could not be quantified. Hospitalization is not the same as new cancer incidence. A headline claiming this study proves cancer causation omits its instability. [Original study abstract](https://pubmed.ncbi.nlm.nih.gov/40881928/)
4. **2026 compilation of case reports:** an Oncotarget review collected 69 publications, predominantly reports covering 333 patients, plus a few population analyses. The existence of case reports can generate hypotheses; without appropriate denominators, comparison groups and ascertainment it cannot establish the frequency or cause of cancer after vaccination. Its inclusion of the Korean/Italian studies does not independently replicate them. [Original review](https://pubmed.ncbi.nlm.nih.gov/41498242/)

A good cancer page needs this evidence hierarchy, rather than merely accumulating favorable experts. The French mortality study is relevant to a claim of mass lethal cancer but is not a comprehensive cancer-incidence study. Findings that vaccination may aid cancer immunotherapy are interesting but not a logical disproof of all proposed adverse effects: different clinical questions need their own evidence. Do not promise that every cancer risk has been ruled out for every future time horizon.

### Residual DNA and SV40

Florida's 3 January 2024 statement called for halting mRNA vaccination on genomic-integration grounds, including inherited effects; September 2024 guidance repeated DNA/immune concerns. These are directly available original policy statements. [January statement](https://www.floridahealth.gov/2024/01/03/florida-state-surgeon-general-calls-for-halt-in-the-use-of-covid-19-mrna-vaccines/) · [September guidance](https://www.floridahealth.gov/2024/09/12/updated-guidance-for-covid-19-boosters-for-the-fall-and-winter-2024-2025-season/)

The correction should admit residual template DNA can be present. A short SV40-derived regulatory sequence is not the complete SV40 virus and does not encode its cancer-causing proteins. Detecting a fragment is not evidence it integrated into recipients' genomes or caused a tumor. The answer should not depend on the overbroad assertion that foreign DNA can never enter a nucleus under any imaginable conditions.

- An independent Slovak analysis tested 15 Comirnaty/Spikevax batches using four complementary methods, finding residual DNA below approved limits and in small fragments. It addresses the sampled batches and measurement claims, not every vial ever made. [npj Vaccines primary paper](https://www.nature.com/articles/s41541-025-01304-9) · [PubMed abstract](https://pubmed.ncbi.nlm.nih.gov/41390528/)
- A separate German laboratory paper explains how mRNA/lipid interference can inflate fluorometric DNA readings without suitable sample preparation. A claimed number of fragments is not itself a toxic dose or integration rate. [Vaccine 2025 primary paper](https://pubmed.ncbi.nlm.nih.gov/40120438/)

**Fact-check quality warning:** the October 2026 AFP cancer article includes an expert's sweeping statement that no peer-reviewed publications show excess cancer among vaccine recipients. The Korean and Italian papers above do report positive associations in some analyses. Do not repeat that absolute formulation. The supported conclusion is that those observations do not establish a causal link, not that no such observations were published.

## 4. Fertility, pregnancy, birth rates and infant outcomes

### Current claim examples

On 14 April 2026 Nicolas Hulscher attributed supposed very high excess infant deaths to earlier vaccination of young women. AFP checked the denominator and actual US mortality change rather than assuming temporal correlation was causal. [Dated post and analysis](https://factcheck.afp.com/doc.afp.com.A88D6G7)

In August 2026 newly released 2021 Fauci messages were presented as proof officials knew vaccination caused miscarriages; the old “82%” calculation also returned. The original conversation discussed a theoretical concern, not an observed 82% causal risk. That percentage excludes ongoing pregnancies from the relevant denominator. AFP documents named political/influencer amplifiers in English, with a French version; no total audience estimate was supplied. [AFP, 13 August 2026](https://factcheck.afp.com/doc.afp.com.C4HE6ZQ) · [French version](https://factuel.afp.com/doc.afp.com.C4MV6CQ) · [Original Thorp testimony used to support the claim](https://www.hsgac.senate.gov/wp-content/uploads/Thorp-Testimony.pdf)

### Evidence and honest boundaries

Suggested lead: **Vaccination can protect during pregnancy. Studies of mRNA COVID vaccines have not found the infertility or miscarriage increases claimed in viral posts. Temporary changes to periods are a different question, and they can happen.**

- **Benefit evidence:** the Canadian CANCOVID-Preg study of 19,899 SARS-CoV-2-affected pregnancies found vaccination associated with less maternal hospitalization, critical-care admission and preterm birth in both Delta and Omicron periods. It is observational, and these historical infected-patient results do not supply the absolute benefit of a particular 2026 dose. [Primary study, published online December 2025](https://jamanetwork.com/journals/jama/fullarticle/2842835)
- **Miscarriage specifically:** a Scottish matched cohort included 18,780 women vaccinated around conception/early pregnancy and found no increased miscarriage odds with either mRNA product. An AstraZeneca signal appeared against historical but not contemporary controls, with substantial clinical-risk differences; it should not be presented as a proven causal effect. Early losses not brought to medical attention and residual confounding remain limitations. [Calvert et al., Nature Communications 2022](https://www.nature.com/articles/s41467-022-33937-y)
- **Trying to conceive:** PRESTO followed 2,126 US/Canadian participants and partners. Vaccination was not appreciably associated with per-cycle conception probability: female FR 1.08 (0.95–1.23), male FR 0.95 (0.83–1.10). Self-reported exposures and selected pregnancy planners limit generalization. [Wesselink et al. primary study](https://pubmed.ncbi.nlm.nih.gov/35051292/)
- **Birth-rate claim:** a 2026 Swedish study of 59,773 women aged 18–45 used time-varying vaccination and adjusted models, finding no decrease in childbirth (adjusted HR 1.03, 0.97–1.09) or increase in recorded miscarriages. Conception timing is inferred and early losses may be missed; national birth-rate declines also reflect behavior and population structure. [Communications Medicine, 21 January 2026](https://www.nature.com/articles/s43856-026-01396-x)
- **Birth defects:** a Nordic registry study of 343,066 liveborn singleton infants found no association of first-trimester vaccination with major congenital anomalies. Because it is a live-birth study, do not use it alone to prove the absence of miscarriage risk. [BMJ 2024](https://www.bmj.com/content/386/bmj-2024-079364)
- **Evidence synthesis:** the 2024 National Academies assessment favored rejection of a causal relationship between Pfizer/Moderna mRNA vaccines and female infertility. Its conclusions are product/outcome-specific and do not imply equal evidence for every platform. [Assessment chapter](https://www.ncbi.nlm.nih.gov/books/NBK607379/)
- **Contrary findings that should not be hidden:** a 2025 Chinese prospective, single-center IVF-FET cohort of 2,733 women reported lower clinical pregnancy/live birth after inactivated vaccines, especially with vaccination near ovarian stimulation (adjusted live-birth RR 0.89, 0.82–0.98). This is neither mRNA evidence nor proof of permanent infertility. Other inactivated-vaccine IVF cohorts found no clear adverse association; clinic selection, treatment timing and confounding matter. Reproductive specialists should assess implications for ART timing rather than a general-audience site inventing a waiting interval. [2025 positive association](https://pubmed.ncbi.nlm.nih.gov/40642511/) · [2022 comparison without significant adverse finding](https://pubmed.ncbi.nlm.nih.gov/35876815/)
- **Menstrual changes:** prospective tracking supports small, usually temporary cycle changes after vaccination. A person reporting a change is not thereby spreading misinformation; the unsupported inference is that this demonstrates sterility. [Primary cycle-timing study](https://pmc.ncbi.nlm.nih.gov/articles/PMC11328953/)

Pregnancy recommendations and eligibility differ by country and change over time. This research map does not provide a current personalized recommendation. A published page should date the guidance separately from the underlying evidence and encourage discussion of individual risk with the person's clinician, without implying that official disagreement creates scientific proof of harm.

## 5. Immune damage, IgG4, VAIDS and persistent symptoms

The documented family includes generalized immune suppression, a supposed vaccine-acquired AIDS syndrome, cancer immune tolerance, and long-term spike poisoning. An August 2026 German fact-check documented renewed claims that a Yale study proved VAIDS. In Malaysia, a February 2026 Malay-language post combined Sucharit Bhakdi's dormant-TB prediction with a false claim that Pfizer had deliberately put TB bacteria in the vaccine. Those are different mechanisms, not interchangeable versions of a single study. [German current recurrence](https://www.mimikama.org/yale-v-aids-durch-covid-impfstoffe/) · [AFP Malaysia](https://factcheck.afp.com/doc.afp.com.98V79T6)

Suggested lead: **COVID vaccines train an immune response. Researchers have found changes in some antibodies after repeated mRNA doses, but that does not establish AIDS, widespread immune collapse or cancer caused by vaccination.**

### Do not overcorrect the science

1. A German study really did find increasing **spike-specific IgG4**, with reduced measured effector functions after repeated mRNA vaccination. It is not evidence that all immunity shuts down. [Irrgang et al., Science Immunology 2023](https://pubmed.ncbi.nlm.nih.gov/36548397/)
2. A **2025 Spanish health-care-worker cohort** found higher IgG4 associated with breakthrough infection, HR 1.8 per tenfold increase (95% CI 1.2–2.7). This is a correlational biomarker result among vaccinated people, not randomized evidence that an extra dose causes more illness than no dose. It did not demonstrate cancer or AIDS. [Journal of Infection primary study](https://www.sciencedirect.com/science/article/pii/S0163445325000672)
3. An **August 2026 study** followed antibody patterns across six doses in 104 Swedish health-care workers, with a small Singapore pediatric cohort. In a 41-person initially infection-naive adult subgroup, the IgG4 response was associated with breakthrough infection. This is an antibody-marker association, not a risk per vaccine dose. The authors emphasize functional mechanisms cannot be established from these data and explicitly do not conclude mRNA safety/efficacy is overturned. Prior infection order matters; results are not automatically applicable to today's largely previously infected population. [Scientific Reports 2026](https://www.nature.com/articles/s41598-026-64190-8) · [Primary abstract](https://pubmed.ncbi.nlm.nih.gov/42618711/)
4. A Yale **exploratory preprint** compared 42 people reporting chronic post-vaccination symptoms with 22 vaccinated controls and found immune/antigen differences needing validation. That selected, small study cannot estimate population incidence, prove causality for all symptoms or establish AIDS. The authors' institutional account stresses preliminary status. Do not dismiss patients' symptoms as imaginary. [Original preprint](https://www.medrxiv.org/content/10.1101/2025.02.18.25322379v1.full) · [Yale explanation](https://news.yale.edu/2025/02/19/immune-markers-post-vaccination-syndrome-indicate-future-research-directions)
5. **Frameshifting is real laboratory biology:** the Nature study found off-target immune responses following modified mRNA translation, but reported no demonstrated adverse human outcome from those products. It proposes sequence-design improvements, not a conclusion that existing vaccines cause mass toxicity. [Mulroney et al., Nature](https://www.nature.com/articles/s41586-023-06800-3)

**Editorial update:** “There is no clinical association whatsoever from IgG4 switching” is now too broad. The correct distinction is between limited evidence of an association with breakthrough infection and unsupported claims about generalized immune failure, cancer or inevitable harm. This should be a visible example of updating the explanation as evidence changes.

Do not assert all vaccine material disappears everywhere within a fixed couple of days, or that it never leaves the injection site. Duration, anatomical compartment, assay and whether material is intact/functional matter. Conversely, detecting antigen is not proof of indefinite increasing production, transmissible shedding or a need for commercial detox products. A standalone detox page would need its own demonstrated reach and treatment-evidence review.

## 6. Effectiveness and natural immunity

HHS's 5 August 2025 original announcement said it would terminate/de-scope 22 mRNA development investments and argued the technology failed to protect effectively against upper-respiratory infection. This is direct evidence of consequential official amplification. It is not evidence that there is no protective effect against hospitalization or death, and COVID findings cannot simply determine efficacy of every future flu product. [HHS original statement](https://www.hhs.gov/press-room/hhs-winds-down-mrna-development-under-barda.html)

Suggested lead: **Updated COVID vaccines can add protection against serious illness even after earlier vaccination or infection. Protection against catching or passing on the virus is less complete and fades.**

- A June 2026 VISION test-negative study used September–December 2025 encounters: 85,725 ED/urgent-care encounters and 26,073 hospitalizations. Estimated 2025–26 vaccine effectiveness was 50% (42–57%) against COVID-associated ED/urgent-care encounters and 55% (41–66%) against hospitalization, with median time since dose about 46–47 days. The comparator was no **current-season** dose, not never vaccinated. Participants were immunocompetent adults; this is not a 12-month, pediatric, product-specific or randomized estimate. [JAMA Network Open primary paper](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2850668)
- The 2024–25 VA study used an active comparator of flu-vaccinated veterans, reducing some health-seeking confounding. It observed estimated effectiveness of 39% against hospitalization and 64% against death. It remains observational and primarily reflects an older, male population; its relative benefits must not be sold as a young healthy adult's absolute risk reduction. [NEJM primary paper](https://www.nejm.org/doi/full/10.1056/NEJMoa2510226)

A good explanation distinguishes infection, symptomatic disease, medically attended disease, hospitalization, death and long COVID. It also admits immune escape and waning, recognizes infection-acquired immunity, and asks how much incremental protection a current dose adds for the person's actual circumstances. “You can still get infected” does not answer that question. A non-significant or negative observational infection-effectiveness estimate should be evaluated for prior-infection, testing and exposure differences, not denied or automatically interpreted as immune damage. The existing COVID/guidance pages can carry this explanation.

## 7. Writing and information architecture

Use benefit-first, direct pages with a short answer, the most relevant actual risk, a small set of strong studies, and clearly separated uncertainty. Specific recommendations:

1. New page: **COVID vaccines and cancer: what the evidence shows.** Lead with severe-COVID benefit; answer the causal cancer question; explain why case reports and the Korean/Italian signals do not prove it; include DNA/SV40 and IgG4 sub-sections. Provide a compact evidence table with product, population, outcome, time window, result and limitation.
2. New page candidate: **COVID vaccination, fertility and pregnancy.** Keep conception, menstrual changes, pregnancy loss, birth defects and infant outcomes separate. Give the positive inactivated-vaccine IVF finding a fair paragraph, not a buried footnote.
3. Update existing deaths page with the 2025 French cohort and 2026 released FDA memo. Preserve existing clotting-risk material and explain historical product relevance.
4. Update existing myocarditis page with “what is known / what remains uncertain,” including MRI follow-up and the young-male Moderna exception to simple infection comparisons.
5. Existing harms page: link to the above instead of reprinting full narratives. Add the biomarker-versus-clinical-disease distinction and a respectful sentence about people with persistent symptoms.
6. Existing amplification page: show dated examples of false authority, numerator/denominator distortion, laboratory-to-clinical leaps and misleading reuse of papers. Do not imply identifiable people knowingly lied unless independently evidenced.
7. Existing guidance page: make current jurisdiction/date visible; do not treat changing public-health policy as a substitute for study appraisal.

Avoid pages titled as sensational claims unless search/navigation needs warrant a clearly corrective subtitle. Avoid repeating the same myth in headings, summaries and social snippets. Show the strongest counterargument in its real form, then explain what changes the inference. Do not respond to “no certainty of zero risk” by pretending that all possibilities deserve equal probability.

## Publication checks and research watch

Before a page is published, a suitably qualified scientific/medical reviewer should verify all current clinical guidance and the specific risk figures. Recommended watch triggers are evidence-driven, not a generic recurring campaign:

- Editorial resolution, reanalysis or independent replication of the Korean cancer paper.
- Larger/independently replicated IgG4 studies with clinical severity outcomes and relevant prior-immunity strata.
- Peer-reviewed validation of persistent-symptom findings and better incidence estimates.
- Longer-term myocarditis follow-up, particularly clinical meaning of persistent MRI abnormalities.
- Further product-specific ART evidence rather than blanket inferences about fertility.
- Any new national safety assessment that changes the possible/probable pediatric-death classification or provides population-level risk estimates.

No external monitoring or automation was created by this research task.

## Source quality and access notes

- Prefer the original scientific paper, denominator, supplement and correction notice over a fact-checker's verdict. Scientific papers cited here range from national linked cohorts to small selected cohorts, laboratory studies and a preprint; they should not be presented as interchangeable evidence.
- Official documents prove what officials stated; institutional status does not guarantee the claim is correct. Manufacturer statements were not used as the sole scientific basis for reassurance.
- AFP, Full Fact and FactCheck.org provide dated claim examples and sometimes archived originals. Many source posts/Perma previews were not directly retrievable, so the attributed content is explicitly based on those outlets' documented investigations where relevant. The original Senate/ACIP/Florida/HHS documents and the released FDA report were accessible directly.
- Several Nature, PubMed and DOI pages intermittently returned sparse/error responses. Equivalent indexed primary abstracts, PMC text or the study team's primary summary were used when available. The French long-term mortality result was checked through its indexed JAMA abstract and French research-team page; the DOI redirect itself did not render reliably. No claim is made to have independently reanalyzed individual-level data.
- The Korean cancer paper's editorial warning was directly checked on its current publisher page. Its access/Altmetric numbers are an access-time snapshot, not validated human reach.
- No paywalled article text, inaccessible original social-media metrics or private group traffic was invented. No assertion of motive is needed to identify the misleading inference.