Studies
Vaccinated vs unvaccinated studies
Studies that compare the health of vaccinated and unvaccinated children, or of children with more and fewer vaccines, focusing on chronic, allergic and autoimmune conditions. They differ greatly in size, design and publication status, so each row gives both. All are observational: they show associations, which can come from the vaccines or from differences between the families who choose them.
At a glance
| Study | Size | Main finding | Status and quality |
|---|---|---|---|
| Henry Ford Health draft (Lamerato, Zervos et al., c.2020), Michigan | 18,468 children; 1,957 unvaccinated | Vaccinated: 2.5 times the rate of any chronic condition; asthma 4.3 times, atopic disease 3.0, autoimmune 6.0 (only 2 cases among the unvaccinated). No difference for cancer, food allergy or autism. | Unpublished draft, never peer-reviewed. Henry Ford Health says it was abandoned for flawed data and methods. Entered into the Senate record on 9 Sep 2025. source source |
| Mawson et al. 2017, homeschooled children, 4 US states | 666 children; 261 unvaccinated | Vaccinated: more reported allergies and neurodevelopmental disorders, but less chickenpox and whooping cough. | A Frontiers journal withdrew an earlier version before publication (2016). Reported retracted by J Transl Sci in May 2017, then reposted; no retraction notice now. source source |
| Hooker & Miller 2020, 3 US practices | Children born 2005-2015 (records) | Vaccinated in year 1: higher unadjusted odds of asthma (4.5), developmental delay (2.2), ear infections (2.1). | Published (SAGE Open Medicine). Expression of concern posted 18 May 2026 while the journal investigated reader concerns. On 28 Aug 2026 SAGE replaced it with a joint publisher–author corrigendum (adds an acknowledgement of the group that referred the practices) and a publisher's note: one concern resolved; whether the data were reused in later articles without disclosure was left inconclusive. Not retracted. expression of concern corrigendum source source |
| Hooker & Miller 2021, survey of families at 3 US practices | Not checked | Vaccinated: higher odds of asthma (17.6), severe allergies (4.3), digestive disorders (13.8), among others. | Published (J Transl Sci, the same publisher as Mawson 2017). Parent-reported, with chart review. source |
| Lyons-Weiler & Thomas 2020, one Oregon practice | 3,324 children; 561 unvaccinated | More office visits for asthma, allergies and other conditions among the vaccinated. | Retracted on 22 Jul 2021 ('conclusions were not supported by strong scientific data'); the authors disagreed. source source |
| KiGGS (Schmitz 2011), Germany national survey | 13,453 children; 0.7% unvaccinated | Atopic disease was not linked to vaccination status (ages 1-5: 12.6% unvaccinated vs 15.0% vaccinated). | Peer-reviewed. Very small unvaccinated group; parent-reported. source |
| Kemp 1997, Christchurch (New Zealand) birth cohort | 1,265 children; 23 unvaccinated | None of the 23 unvaccinated had asthma before age 10, vs 23.1% of vaccinated children. | Peer-reviewed. Only 23 unvaccinated; the authors could not exclude differences in health-service use. source |
| McKeever 2004, UK GP records birth cohort | Large GP database | The link between vaccination and asthma or eczema appeared only in children with the fewest GP visits. | Peer-reviewed. Authors: explained by under-diagnosis. Critics: this does not rule out an effect. source source |
| Hviid 2004, Denmark (all births 1990-2000) | 681 cases of type 1 diabetes in 4.7 million person-years | No link between any childhood vaccine and type 1 diabetes (e.g. MMR rate ratio 1.14, 95% CI 0.90-1.45). | Peer-reviewed (NEJM). source |
| Andersson 2025, Denmark (vaccine aluminium dose) | 1,224,176 children | No rise in any of 50 autoimmune, allergic or neurodevelopmental conditions with more aluminium (autoimmune HR 0.98 per mg). | Peer-reviewed (Annals of Internal Medicine). Compares amounts of vaccine aluminium, not vaccinated vs unvaccinated. source |
| Daley 2023, US Vaccine Safety Datalink (aluminium dose) | 326,991 children | Slightly higher persistent asthma with more vaccine aluminium (HR 1.19-1.26 per mg). | Peer-reviewed. The authors note small effects and possible residual confounding; more study called for. source |
| Glanz 2013, US Vaccine Safety Datalink (undervaccination) | 323,247 children | Undervaccinated children had fewer outpatient visits (IRR 0.89) but more hospital admissions (IRR 1.21). | Peer-reviewed. Shows that care-seeking differs by vaccination status. source |
| Miller & Goldman 2011, 34 countries | 34 countries | Countries giving more infant vaccine doses had higher infant mortality (r = 0.70). | Peer-reviewed. Country-level comparison, so it cannot show effects on individuals; it does not cover chronic illness. source |
The Henry Ford Health draft (c.2020)
A retrospective study of 18,468 children born 2000-2016 in a Michigan health plan, by Lois Lamerato, Abigail Chatfield, Amy Tang and Marcus Zervos. It compared children with no vaccines (1,957) against those with one or more (16,511), using medical records and the Michigan immunisation registry. The draft was entered into the record of a US Senate hearing on 9 September 2025. source source
| Outcome (vaccinated vs unvaccinated) | Adjusted hazard ratio (95% CI), Table 3 | Cases (vaccinated vs unvaccinated) |
|---|---|---|
| Any chronic condition (composite) | 2.54 (2.16-2.97) | 4,732 vs 160 cases |
| Asthma | 4.29 (3.26-5.65) | 2,867 vs 52 |
| Atopic disease | 3.03 (2.01-4.57) | 946 vs 23 |
| Autoimmune disease | 5.96 (1.48-24.11) | 201 vs 2 |
| Eczema (not in the composite) | 1.31 (1.13-1.52) | - |
| Neurodevelopmental disorder | 5.53 (2.91-10.51) | - |
| Cancer (used as a control outcome) | 0.90 (0.51-1.59), no difference | 169 vs 13 |
| Food allergy | 1.40 (0.97-2.02), no clear difference | 577 vs 30 |
| Diabetes | not calculable (0 unvaccinated cases) | 42 vs 0 |
A hazard ratio of 2 means about twice the rate. Ten years in, 43% of vaccinated and 83% of unvaccinated children had no chronic condition recorded. source
Methodological criticisms and the draft's responses
- Follow-up time: The median follow-up was 461 days for unvaccinated children and 970 for vaccinated. Henry Ford says 75% of the unvaccinated were observed only up to age 3, before many chronic conditions can be diagnosed. source source
- Healthcare use (detection bias): Vaccinated children averaged 7 healthcare encounters a year and unvaccinated children 2 (almost 5 if they had a chronic condition). More visits give more chances to record a diagnosis. The draft re-ran the analysis using only children with at least one encounter (HR 1.87) and with longer enrolment (1, 3 and 5 years: HR 2.84-4.05); the association remained. Other studies show care-seeking differs by vaccination status. source source source
- Group differences: Henry Ford says the unvaccinated group had more boys, more white children, and less prematurity and breathing problems at birth. The draft lacked data on income, diet and lifestyle. source source
- Exposure definition: Any vaccine vs none. The number of vaccines and the timing between vaccine and diagnosis were not analysed. source
- Small numbers for some outcomes: The autoimmune result rests on 2 cases among the unvaccinated, hence the wide confidence interval (1.48-24.11). The abstract and Table 3 give slightly different hazard ratios (e.g. autoimmune 4.79 vs 5.96). source
What each side says
- Aaron Siri (lawyer for ICAN), Senate testimony: ICAN asked Dr Zervos in 2017 to run the study; the ICAN film preview played at the hearing says he promised to publish whatever the results. Siri says the authors told him they had not submitted it because they feared for their jobs and did not want to make doctors uncomfortable, and he calls this an example of 'the corruption of science'. Responding to the follow-up-time and detection-bias criticisms, he pointed to the draft's sensitivity analyses and its use of cancer as a control outcome. source source
- Henry Ford Health statements (Sep 2025): The draft was 'immediately shelved upon the first internal peer review because of serious issues with its data and methodology' and was never submitted. The health system denies suppressing it for its results, says the draft was shared without its knowledge, and sent a cease-and-desist letter to the film's makers. source source
- Dr Jake Scott (minority witness), Senate testimony: Called the draft 'fundamentally flawed by design': vaccinated children had twice the follow-up time and more healthcare visits, which he called 'classic detection bias'. source source
- Film 'An Inconvenient Study' (ICAN / Del Bigtree): A documentary about the draft, premiered online on 12 Oct 2025 (originally announced for 3 Oct). A 5-minute preview was played at the hearing. source
The Senate hearing, 9 September 2025
US Senate Committee on Homeland Security and Governmental Affairs, Permanent Subcommittee on Investigations (Chairman Ron Johnson, Ranking Member Richard Blumenthal): "How the Corruption of Science has Impacted Public Perception and Policies Regarding Vaccines", Hart Senate Office Building SH-216, 2:00 pm. source source
| Witness | Role | Called by | Introduced at |
|---|---|---|---|
| Aaron Siri | Managing partner, Siri & Glimstad LLP; lawyer for ICAN | Majority (Chairman Johnson) | 0:22:32 |
| Toby Rogers, PhD | Fellow, Brownstone Institute | Majority (Chairman Johnson) | 0:29:11 |
| Jake Scott, MD | Infectious disease specialist, Stanford University School of Medicine | Minority (Ranking Member Blumenthal) | 0:34:51 |
| Robert W. Malone, MD | Member, CDC Advisory Committee on Immunization Practices (testified in a personal capacity); panel II | Sworn in by the Chairman after being mentioned during the hearing | 3:01:59 |
Key moments (links open the full video at that time)
Timestamps are for the full-length NTD stream on YouTube and were checked against its captions. The official Senate video is on the committee page. source source
- 0:00:06 Chairman Ron Johnson opens the hearing and introduces the Henry Ford draft.
- 0:05:41 Ranking Member Richard Blumenthal's opening statement.
- 0:16:36 Witnesses sworn in; a 5-minute ICAN preview of the film is played (from 0:16:59).
- 0:22:32 Aaron Siri introduced; his opening statement on the IOM 2013 report and the Henry Ford draft starts at 0:23:00.
- 0:29:11 Dr Toby Rogers introduced; his opening statement starts at 0:29:37.
- 0:34:51 Dr Jake Scott introduced; his opening statement starts at about 0:35:10.
- 0:37:53 Dr Scott: the Henry Ford draft is 'fundamentally flawed by design' (follow-up time and detection bias).
- 0:41:09 Siri, asked to defend the draft, describes its sensitivity analyses.
- 1:20:47 Ranking Member Blumenthal asks Siri why the study was not made public sooner; Siri replies that the authors would not submit it.
- 3:01:59 Hearing reconvenes; Dr Robert Malone is sworn in and gives a statement (from 3:02:21).
Why aren't more of these studies done?
The ethics argument
A randomised trial that withholds proven vaccines from some children would expose them to preventable disease and weaken community protection. The IOM (2013) recommended that HHS 'refrain from initiating randomized controlled trials' comparing fully vaccinated with unvaccinated children. source
The counter-argument
Children whose parents have already chosen not to vaccinate exist, so observational studies of them are ethical. The IOM found that 'no studies have compared the differences in health outcomes… between entirely unimmunized populations of children and fully immunized children', and called analyses of the Vaccine Safety Datalink 'the most feasible approach'. A CDC-linked white paper (2016) set out how to do this. source source
Practical problems
Fewer than 1% of US children receive no vaccines, so large samples are hard to assemble. Unvaccinated families differ in income, lifestyle and healthcare use, which can bias the results either way. The IOM said such comparisons 'cannot be fully accounted for' and religious groups are too small to give firm answers. source source
What has been done
Registry studies compare specific vaccines, doses or aluminium amounts (Denmark, the Vaccine Safety Datalink). A few studies include a small fully unvaccinated group (Kemp, KiGGS). Studies with large unvaccinated groups are either unpublished (Henry Ford), retracted (Lyons-Weiler) or survey-based with selection problems (Mawson, Hooker & Miller). The IOM concluded there was 'no evidence that the schedule is not safe', while saying long-term studies of the whole schedule 'have not been conducted'. source source source
More studies: vaccinated vs unvaccinated, and placebo-controlled trials
Worth knowing
No trial has ever compared fully vaccinated with fully unvaccinated children, blinded, over years. In the US, vaccine makers are largely shielded from injury lawsuits under a 1986 law, and injury claims are paid by a government scheme instead. In the UK, the government pays a one-off Vaccine Damage Payment.
Fine print from the sources: the HRSA page says the US programme is a no-fault alternative to the courts, and a claimant who rejects its decision may still sue the vaccine company or provider; the statute bars suits for side effects that were unavoidable despite proper preparation, directions and warnings 42 U.S.C. 300aa-22. GOV.UK says the UK payment is ‘not a compensation scheme’ and that legal action is still possible.
53 further studies, each linked to its primary source (DOI, PubMed or the journal). Where an entry has a status other than ordinary peer review it says so in a coloured label. Studies are grouped by design. We include results that point in different directions, and say what each study could not show. Results are summarised from each paper’s abstract or trial record, in plain words; the numbers are the authors’ own. This guide takes no position on what they mean.
Jump to: Vaccinated vs unvaccinated · Studies finding no difference · Placebo-controlled trials · Retracted or withdrawn
Vaccinated vs unvaccinated, and more vs fewer vaccines (15)
Observational studies and reviews. They show associations, which can come from the vaccines or from differences between the families who choose them. Two (Steiner schools) compare lifestyles, not only vaccination; the last two are reviews. Studies that found no difference are in the next group.
| Study | What was compared | Plain finding | Source |
|---|---|---|---|
| Vaccination and neurodevelopmental disorders: a study of nine-year-old children enrolled in Medicaid Mawson AR, Jacob B (2025). Science, Public Health Policy and the Law 6:2019-2025 Journal-published; not in PubMed | Florida Medicaid claims; 47,155 children, of whom 5,123 (10.9%) had no vaccination claims | 27.8% of vaccinated vs 11% of 'unvaccinated' children had at least one neurodevelopmental diagnosis (odds ratio 3.12). 'Unvaccinated' was inferred from having no vaccine billing codes; funded by the National Vaccine Information Center. | paper |
| Preterm birth, vaccination and neurodevelopmental disorders: a cross-sectional study of 6- to 12-year-old vaccinated and unvaccinated children Mawson AR, Bhuiyan AZ, Jacob B, Ray BD (2017). Journal of Translational Science 3(3):1-8 Journal-published; not in PubMed | Online survey of mothers; 666 homeschooled children, 261 unvaccinated | Vaccination was linked to neurodevelopmental disorders in children born at term (odds ratio 2.7); the link was stronger in vaccinated children born preterm. The authors call it a pilot study that needs confirming. | paper |
| Health versus disorder, disease, and death: unvaccinated persons are incommensurably healthier than vaccinated Garner J (2022). International Journal of Vaccine Theory, Practice, and Research 2(2):670-686 Journal-published; not in PubMed | Self-reported survey of unvaccinated people (210 adults, 1,272 children), compared with general US population figures | The unvaccinated sample reported lower rates of chronic conditions than the general population. The comparison is self-reported and not matched. | paper |
| Comparative health outcomes of vaccinated versus unvaccinated children: a retrospective cohort analysis Yengst T, Ray H, Jablonowski K, Hooker BS (2026). Zenodo preprint (26 Sep 2026) Preprint, not peer-reviewed | Records from 8 US paediatric practices; 6,239 children (2,414 vaccinated before 13 months, 3,825 not) | Higher odds of rhinitis (1.8), ear infection (2.0), respiratory disease (1.6) and strep (2.3) in the vaccinated group; developmental delay was not significantly different overall. The authors are affiliated with Children's Health Defense. | paper |
| Effects of diphtheria-tetanus-pertussis or tetanus vaccination on allergies and allergy-related respiratory symptoms among children and adolescents in the United States Hurwitz EL, Morgenstern H (2000). J Manipulative Physiol Ther 23(2):81-90 Peer-reviewed | US national survey (NHANES III), parent recall; vaccinated vs unvaccinated; few unvaccinated children | History of asthma: odds ratio 2.00 in the vaccinated, but the range (0.59 to 6.74) includes no difference. Allergy-related breathing symptoms: odds ratio 1.63 (1.05 to 2.54). The authors say the small unvaccinated group limits firm conclusions. | paper |
| The relationship between vaccine refusal and self-report of atopic disease in children Enriquez R, Addington W, Davis F, et al. (2005). J Allergy Clin Immunol 115(4):737-44 Peer-reviewed | Mailed survey of National Vaccine Information Center member households; 515 never vaccinated, 423 partly, 239 fully vaccinated children | Parents who refused vaccines reported less asthma, hay fever and eczema in their children. The authors say unmeasured lifestyle differences or bias could explain this. | paper |
| Atopy in children of families with an anthroposophic lifestyle Alm JS, Swartz J, Lilja G, Scheynius A, Pershagen G (1999). Lancet 353(9163):1485-8 Peer-reviewed | Swedish children aged 5-13 at Steiner (anthroposophic) schools vs standard schools. Lower vaccine use is one lifestyle difference; this is not a pure vaccinated-vs-unvaccinated comparison | Less atopy (allergy-type disease) in the Steiner-school children (odds ratio 0.62). The authors point to the whole lifestyle, not vaccination alone. | paper |
| Allergic disease and sensitization in Steiner school children Flöistrup H, Swartz J, Bergström A, et al. (2006). J Allergy Clin Immunol 117(1):59-66 Peer-reviewed | Children at Steiner and comparison schools in 5 European countries. Not a pure vaccinated-vs-unvaccinated comparison | Less allergic disease with restricted antibiotic and fever-medicine use; MMR vaccination was linked to more rhinoconjunctivitis, and measles infection to less IgE-mediated eczema. | paper |
| Transient suppression of atopy in early childhood is associated with high vaccination coverage Grüber C, Illi S, Lau S, et al. (2003). Pediatrics 111(3):e282-8 Peer-reviewed | German birth cohort (943 followed to age 5), grouped by number of vaccine doses received. A finding in favour of vaccination, included for balance | More vaccine doses went with less atopic dermatitis and asthma in early childhood (atopic dermatitis at age 3: 27.6% in the lowest-dose group vs 4.5% in the highest). | paper |
| Early childhood infection and atopic disorder Farooqi IS, Hopkin JM (1998). Thorax 53(11):927-32 Peer-reviewed | Records of 1,934 children born 1975-84 at one family-doctor practice in Oxfordshire (UK) | Whole-cell pertussis (whooping cough) immunisation predicted later asthma, hay fever or eczema (odds ratio 1.76, range 1.39 to 2.23), as did maternal atopy and early antibiotics. The authors urge caution because of confounding and reverse causation. | paper |
| Delay in diphtheria, pertussis, tetanus vaccination is associated with a reduced risk of childhood asthma McDonald KL, Huq SI, Lix LM, Becker AB, Kozyrskyj AL (2008). J Allergy Clin Immunol 121(3):626-31 Peer-reviewed | 11,531 Manitoba (Canada) children with at least 4 DPT doses; compares timing, not vaccinated vs unvaccinated | Asthma risk was about halved when the first DPT dose was delayed by more than 2 months. | paper |
| Routine vaccinations and child survival: follow up study in Guinea-Bissau, West Africa Kristensen I, Aaby P, Jensen H (2000). BMJ 321(7274):1435-8 Peer-reviewed | Follow-up of 15,351 women and their children in rural Guinea-Bissau; recipients vs non-recipients of each vaccine | Children who had had one dose of DTP had higher mortality than those with none (ratio 1.84); BCG (0.55) and measles vaccine (0.48) went with lower mortality. | paper |
| The introduction of diphtheria-tetanus-pertussis and oral polio vaccine among young infants in an urban African community: a natural experiment Mogensen SW, Andersen A, Rodrigues A, Benn CS, Aaby P (2017). EBioMedicine 17:192-198 Peer-reviewed | Children in Bissau, Guinea-Bissau, vaccinated earlier or later between 3 and 5 months by birthday | Among 3-5-month-olds, those who had received DTP had a mortality hazard ratio of 5.00 (1.53 to 16.3) compared with not-yet-vaccinated children. The authors conclude DTP was associated with higher mortality. | paper |
| Testing the hypothesis that diphtheria-tetanus-pertussis vaccine has negative non-specific and sex-differential effects on child survival in high-mortality countries Aaby P, Benn C, Nielsen J, et al. (2012). BMJ Open 2(3):e000707 Peer-reviewed | Review of observational studies and two trials in high-mortality countries | The authors conclude DTP is associated with higher mortality in girls in these settings and call for randomised trials of DTP. | paper |
| Association of BCG, DTP, and measles containing vaccines with childhood mortality: systematic review Higgins JPT, Soares-Weiser K, López-López JA, et al. (2016). BMJ 355:i5170 Peer-reviewed | Systematic review of randomised and non-randomised studies in low-income settings | BCG and measles vaccines were associated with lower overall mortality. DTP was linked to a possible rise (relative risk 1.38, range 0.92 to 2.08, from 10 studies at high risk of bias). The authors ask for randomised trials. | paper |
Studies that found no difference, or favoured vaccination (5)
Included for balance. Most compare large groups, some of them whole countries.
| Study | What was compared | Plain finding | Source |
|---|---|---|---|
| Diphtheria, pertussis, poliomyelitis, tetanus, and Haemophilus influenzae type b vaccinations and risk of eczema and recurrent wheeze in the first year of life: the KOALA birth cohort study Kummeling I, Thijs C, Stelma F, et al. (2007). Pediatrics 119(3):e367-73 Peer-reviewed | Dutch birth cohort, 2,545 infants, including 182 never vaccinated | No significant difference in eczema or recurrent wheeze by vaccination status at 6 months. | paper |
| Measles, mumps, rubella vaccination and autism: a nationwide cohort study Hviid A, Hansen JV, Frisch M, Melbye M (2019). Ann Intern Med 170(8):513-520 Peer-reviewed | All 657,461 children born in Denmark 1999-2010; MMR-vaccinated vs unvaccinated | Autism hazard ratio 0.93 (0.85 to 1.02) for MMR-vaccinated vs unvaccinated: no increased risk found. | paper |
| Autism occurrence by MMR vaccine status among US children with older siblings with and without autism Jain A, Marshall J, Buikema A, et al. (2015). JAMA 313(15):1534-40 Peer-reviewed | 95,727 privately insured US children with older siblings; MMR 0, 1 or 2 doses | MMR vaccination was not associated with a higher risk of autism at any age, including in children whose older sibling had autism. A correction (erratum) was published in JAMA in 2016. | paper |
| Safety of thimerosal-containing vaccines: a two-phased study of computerized health maintenance organization databases Verstraeten T, Davis RL, DeStefano F, et al. (2003). Pediatrics 112(5):1039-48 Peer-reviewed | US health-plan databases (124,170 infants in phase 1), compared by amount of thimerosal (a mercury-containing preservative) received; not vaccinated vs unvaccinated | Some links appeared in two health plans (tics in one, language delay in another) but not in the third, and none for autism or attention-deficit disorder; the authors found no consistent significant association. | paper |
| Childhood vaccination and allergy: a systematic review and meta-analysis Navaratna S, Estcourt MJ, Burgess J, et al. (2021). Allergy 76(8):2514-2531 Peer-reviewed | Systematic review and meta-analysis of studies of commonly used childhood vaccines | No evidence that commonly used childhood vaccines raise the later risk of allergic disease; BCG and measles vaccines were linked to less eczema. | paper |
Placebo-controlled trials (27 trials, 3 re-analyses)
In a placebo-controlled trial, volunteers are randomly given the vaccine or a control injection. What the ‘placebo’ contained matters, so the so the last-but-one column sayslsquo;control group receivedso the last-but-one column saysrsquo; column says. Where the paper and trial record do not say, we say that too. Most of these test one vaccine for one disease, not the whole childhood schedule, and several are in adults.
| Study | Trial | Plain finding | What the control group received | Source |
|---|---|---|---|---|
| Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine Polack FP et al. (2020). N Engl J Med 383:2603-15 Peer-reviewed | Adults and teens 16+, 43,548 randomised (Pfizer-funded) | 8 vs 162 Covid-19 cases (vaccine vs placebo). Serious side effects were low and similar in both groups. | Saline (salt water) injection, stated in the paper | paper |
| Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine Baden LR et al. (2021). N Engl J Med 384:403-16 Peer-reviewed | 30,420 adults (Moderna, US government funded) | 94.1% efficacy against Covid-19. Serious side effects were rare and similar in both groups. | Saline (0.9% sodium chloride), per the trial record NCT04470427 | paper |
| Safety and efficacy of single-dose Ad26.COV2.S vaccine against Covid-19 Sadoff J et al. (2021). N Engl J Med 384:2187-201 Peer-reviewed | Adults, ENSEMBLE trial (Janssen) | Protection against symptomatic Covid-19. Serious side effects were balanced; 3 deaths in the vaccine group vs 16 in the placebo group. | Saline (salt water) injection, stated in the paper (the trial record says only 'placebo') | paper |
| Safety and efficacy of NVX-CoV2373 Covid-19 vaccine Heath PT et al. (2021). N Engl J Med 385:1172-83 Peer-reviewed | UK adults (Novavax-funded) | 89.7% efficacy. Serious side effects were low and similar. | Saline, per the trial record NCT04583995 | paper |
| Safety, immunogenicity, and efficacy of the BNT162b2 Covid-19 vaccine in adolescents Frenck RW et al. (2021). N Engl J Med 385:239-50 Peer-reviewed | Ages 12-15 (Pfizer) | 0 Covid-19 cases among vaccinated vs 16 among placebo recipients. The authors report a favourable safety profile. | Saline (salt water) injection, stated in the paper | paper |
| Evaluation of mRNA-1273 SARS-CoV-2 vaccine in adolescents Ali K et al. (2021). N Engl J Med 385:2241-51 Peer-reviewed | Ages 12-17, 2:1 randomisation (Moderna) | 0 Covid-19 cases in the vaccine group vs 4 in the placebo group. | Saline (0.9% sodium chloride), per the trial record NCT04649151 | paper |
| Evaluation of the BNT162b2 Covid-19 vaccine in children 5 to 11 years of age Walter EB et al. (2022). N Engl J Med 386:35-46 Peer-reviewed | Ages 5-11, 2:1 randomisation (Pfizer) | 3 vs 16 Covid-19 cases (efficacy 90.7%). The authors describe the regimen as safe, immunogenic and efficacious. | Saline (salt water) injection, stated in the paper | paper |
| Efficacy of an adjuvanted herpes zoster subunit vaccine in older adults (ZOE-50) Lal H et al. (2015). N Engl J Med 372:2087-96 Peer-reviewed | Adults 50+, about 15,000 per trial record (GSK-funded) | 97.2% efficacy against shingles. Strong short-term reactions (grade 3) were more common with the vaccine: 17.0% vs 3.2%. | Sodium chloride (NaCl) solution, per the trial record NCT01165177 | paper |
| Efficacy of the herpes zoster subunit vaccine in adults 70 years of age or older (ZOE-70) Cunningham AL et al. (2016). N Engl J Med 375:1019-32 Peer-reviewed | Adults 70+ (GSK-funded) | Fewer cases of shingles and of long-lasting nerve pain after shingles. Reactions in the first 7 days: 79.0% vs 29.5%. | Sodium chloride (NaCl) solution, per the trial record NCT01165229 | paper |
| Safety and immunogenicity of Tdap immunization during pregnancy Munoz FM et al. (2014). JAMA 311(17):1760-9 Peer-reviewed | Small trial: 48 pregnant women (33 vaccine, 15 placebo) | No increase in side effects found in the women or their babies; the authors call it a preliminary assessment. Two corrections (errata) were published in JAMA in 2017. | Saline (0.9% sodium chloride) during pregnancy, with Tdap given after delivery, per the trial record NCT00707148 | paper |
| A vaccine to prevent herpes zoster and postherpetic neuralgia in older adults (Shingles Prevention Study) Oxman MN et al. (2005). N Engl J Med 352:2271-84 Peer-reviewed | 38,546 adults 60+ | Shingles burden of illness down 61.1%, long-lasting nerve pain down 66.5%; injection-site reactions were more common with the vaccine. | Contents not stated in the abstract | paper |
| Efficacy and safety of a bivalent RSV prefusion F vaccine in older adults (RENOIR) Walsh EE et al. (2023). N Engl J Med 388:1465-77 Peer-reviewed | Adults 60+, about 39,000 (Pfizer-funded) | Prevented RSV lung illness. Serious side effects were 2.3% in each group. | Contents not stated in the abstract or the trial record | paper |
| Bivalent prefusion F vaccine in pregnancy to prevent RSV illness in infants (MATISSE) Kampmann B et al. (2023). N Engl J Med 388:1451-64 Peer-reviewed | Pregnant women and their babies (Pfizer-funded) | Side effects were similar: 13.8% vs 13.1% in the women, 37.1% vs 34.5% in the babies. | Contents not stated in the abstract or the trial record | paper |
| Safety and efficacy of a pentavalent human-bovine (WC3) reassortant rotavirus vaccine Vesikari T et al. (2006). N Engl J Med 354:23-33 Peer-reviewed | Infants, about 34,000 per group in the safety analysis (Merck-funded) | Bowel blockage (intussusception) 12 vs 15 cases (vaccine vs placebo); less rotavirus diarrhoea. | Oral placebo; contents not stated in the abstract | paper |
| Safety and efficacy of an attenuated vaccine against severe rotavirus gastroenteritis Ruiz-Palacios GM et al. (2006). N Engl J Med 354:11-22 Peer-reviewed | 63,225 infants in Latin America and Finland (GSK-funded) | 85% efficacy against severe rotavirus diarrhoea; intussusception 6 vs 7 cases (vaccine vs placebo). | Contents not stated in the abstract | paper |
| Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand (RV144) Rerks-Ngarm S et al. (2009). N Engl J Med 361:2209-20 Peer-reviewed | 16,402 adults; an HIV vaccine | Modest, mixed result: 31.2% efficacy in one analysis (P=0.04) but not significant in the main intention-to-treat analysis (P=0.08). | Placebo injections: the ALVAC carrier without the virus plus an AIDSVAX placebo, per the trial record NCT00223080 | paper |
| Clinical efficacy and safety of a novel tetravalent dengue vaccine in healthy children in Asia (CYD14) Capeding MR et al. (2014). Lancet 384:1358-65 Peer-reviewed | Children 2-14, 2:1 randomisation (Sanofi Pasteur-funded) | 56.5% efficacy against symptomatic dengue; serious side effects were consistent with those expected in this age group. | Saline (0.9% sodium chloride), per the trial record NCT01373281 | paper |
| Efficacy of a tetravalent dengue vaccine in children in Latin America (CYD15) Villar L et al. (2015). N Engl J Med 372:113-23 Peer-reviewed | Children 9-16, 2:1 randomisation (Sanofi Pasteur-funded) | 64.7% efficacy in the main analysis; the abstract reports safety similar to placebo. | Saline (0.9% sodium chloride), per the trial record NCT01374516 | paper |
| Effect of dengue serostatus on dengue vaccine safety and efficacy Sridhar S et al. (2018). N Engl J Med 379:327-40 Peer-reviewed | Later reanalysis of three placebo-controlled dengue vaccine trials | A later finding: higher risk of severe disease and hospital stays in vaccinated people who had never had dengue; protection in those who had. | Controls in the original trials (the two dengue trials above used saline) | paper |
| Influenza vaccination of pregnant women and protection of their infants Madhi SA et al. (2014). N Engl J Med 371:918-31 Peer-reviewed | Pregnant women in South Africa, with and without HIV | Partial protection against confirmed flu: about 50% efficacy in women without HIV (50.4%) and in their babies (48.8%). | Contents not stated in the abstract | paper |
| The efficacy in Navajo infants of a conjugate vaccine consisting of Haemophilus influenzae type b polysaccharide and Neisseria meningitidis outer-membrane protein complex Santosham M et al. (1991). N Engl J Med 324:1767-72 Peer-reviewed | Navajo infants, about 2,600 per group | Serious Hib infection before age 18 months: 1 case in the vaccine group vs 22 in the placebo group. | Contents not stated in the abstract | paper |
| Efficacy of nine-valent pneumococcal conjugate vaccine against pneumonia and invasive pneumococcal disease in The Gambia Cutts FT et al. (2005). Lancet 365(9465):1139-46 Peer-reviewed | Gambian infants | 50% efficacy against invasive disease (all types) and 16% against deaths; serious side effects 110 vaccine vs 131 placebo. A correction (erratum) was published in Lancet 2005;366:28. | Contents not stated in the abstract | paper |
| Hepatitis B vaccine: demonstration of efficacy in a controlled clinical trial in a high-risk population in the United States Szmuness W et al. (1980). N Engl J Med 303:833-41 Peer-reviewed | High-risk adults, US | Hepatitis B or silent infection in 1.4-3.4% of vaccine recipients vs 18-27% of placebo recipients. | Contents not stated in the abstract | paper |
| Comparative efficacy of inactivated and live attenuated influenza vaccines Monto AS et al. (2009). N Engl J Med 361:1260-7 Peer-reviewed | Healthy adults, 2007-08 flu season | Absolute efficacy against influenza A: 72% for the inactivated vaccine, 29% (range -14 to 55) for the live nasal vaccine. | Contents not stated in the abstract | paper |
| Prevention of symptomatic seasonal influenza in 2005-2006 by inactivated and live attenuated vaccines Ohmit SE et al. (2008). J Infect Dis 198:312-7 Peer-reviewed | Healthy adults | The live nasal vaccine's efficacy was not statistically greater than placebo for one endpoint (8%, range -194 to 67). | Contents not stated in the abstract | paper |
| Quadrivalent vaccine against human papillomavirus to prevent anogenital diseases (FUTURE I) Garland SM et al. (2007). N Engl J Med 356:1928-43 Peer-reviewed | 5,455 women aged 16-24 (Merck-funded) | Vaccine efficacy 100% for each main end point in the per-protocol group. | NOT saline: an aluminium-containing adjuvant (AAHS) 'placebo', as described by Doshi et al. 2020 (below) | paper |
| A placebo-controlled trial of a pertussis-toxoid vaccine Trollfors B et al. (1995). N Engl J Med 333:1045-50 Peer-reviewed | Swedish infants, about 1,725 per group | 71% efficacy of the whooping-cough vaccine. | NOT an inert placebo: the control group received diphtheria and tetanus toxoid vaccine (an active vaccine) | paper |
Re-analyses of placebo-controlled trial data (3)
| Study | What was done | Plain finding | Control in the original trials | Source |
|---|---|---|---|---|
| Adjuvant-containing control arms in pivotal quadrivalent human papillomavirus vaccine trials: restoration of previously unpublished methodology Doshi P, Bourgeois F, Hong K, Jones M (2020). BMJ Evid Based Med 25(6):213 Peer-reviewed | Review of registry entries, papers, consent forms and clinical study reports for five Gardasil trials | The aluminium-containing control was described in all the papers and study reports but in none of the registry entries; in 3 of 5 trials the consent forms called it 'inactive'. The authors say a non-placebo control may have obscured the safety picture. | Aluminium-containing adjuvant (AAHS) | paper |
| Serious adverse events of special interest following mRNA COVID-19 vaccination in randomized trials in adults Fraiman J, Erviti J, Jones M, Greenland S, Whelan P, Kaplan RM, Doshi P (2022). Vaccine 40(40):5798-5805 Peer-reviewed | Re-analysis of the Pfizer and Moderna adult trials (saline placebo) | Combined, 16% higher risk of serious side effects in vaccine recipients (risk ratio 1.16, range 0.97 to 1.39, which includes no difference); for a listed set of 'adverse events of special interest' the ratio was 1.43 (1.07 to 1.92). The authors call for the participant-level data to be released. | Saline, as in the original trials | paper |
| Randomized clinical trials of COVID-19 vaccines: do adenovirus-vector vaccines have beneficial non-specific effects? Benn CS, Schaltz-Buchholzer F, Nielsen S (2023). iScience 26(5):106733 Peer-reviewed | Meta-analysis of placebo-controlled Covid-19 vaccine trials | Overall death rate ratio 1.03 (0.63 to 1.71) for mRNA vaccines and 0.37 (0.19 to 0.70) for adenovirus-vector vaccines vs placebo. The authors note controls were offered the vaccine early after the trials were unblinded. | Placebo, as in the original trials | paper |
Retracted or withdrawn (3)
These compared vaccinated and unvaccinated people. A retraction or withdrawal is a publisher’s decision about the paper’s reliability; the authors may disagree, and where they have said so it is noted. See also Retracted and removed papers above.
| Study | What was compared | What it originally reported | Status and reason | Source |
|---|---|---|---|---|
| Hepatitis B vaccination of male neonates and autism diagnosis, NHIS 1997-2002 Gallagher CM, Goodman MS (2010). J Toxicol Environ Health A 73(24):1665-77 RETRACTED | US national health survey; boys vaccinated as newborns vs never or later vaccinated (31 autism cases) | Originally reported about three times the odds of a parent-reported autism diagnosis in boys vaccinated as newborns. | RETRACTED 21 May 2026 by the journal and publisher after an independent statistical review found fundamental methodological flaws (cross-sectional design, only 31 autism cases, incomplete adjustment for other factors). The authors do not agree with the retraction. | paper retraction notice Retraction Watch |
| Hepatitis B triple series vaccine and developmental disability in US children aged 1-9 years Gallagher C, Goodman M (2008). Toxicol Environ Chem 90(5):997-1008 RETRACTED | US national health survey (NHANES 1999-2000); 1,824 children; among boys, 46 vaccinated vs 7 unvaccinated | Originally reported about nine times the odds of receiving early-intervention or special-education services in vaccinated boys. | RETRACTED 13 Oct 2026 by the journal and publisher. This guide could not read the stated reason; see the notice. | paper retraction notice |
| Increased Mortality Associated with 2-Month Old Infant Vaccinations Jablonowski K, Hooker B (Children's Health Defense) (2025). Preprints.org 202512.1865 (posted 22 Dec 2025) WITHDRAWN | Louisiana death records 2013-2024; 1,775 children who died before age 3 | The authors concluded children vaccinated in month 2 were more likely to die in month 3 than those not vaccinated then. | WITHDRAWN by Preprints.org on 14 Jan 2026 on its advisory board's advice. The publisher said this was not a finding of misconduct and cited methodological concerns. The authors say the board had no vaccine-safety experts, and reposted it elsewhere. | paper Retraction Watch |
Some publisher sites (Taylor & Francis, SAGE, Preprints.org, Zenodo) block automated checks; their DOIs were confirmed in the Crossref or DataCite registries and the notices are linked.
Retracted and removed papers
Removal alone does not prove a paper right or wrong; here is each case so you can weigh it.
Vaccine-safety papers that a journal or preprint server retracted, withdrew or removed, with the journal's stated reason. Only cases with a notice or a publisher statement this guide could check are listed. Listed by publication date.
Wakefield et al., The Lancet
- Main claim
- 12 children with bowel disease and developmental regression; in 8, parents linked the start of symptoms to MMR vaccination.
- Journal's stated reason
- After the UK General Medical Council ruling: claims that children were 'consecutively referred' and that investigations were ethics-approved were 'proven to be false'.
- Author response
- Author response: no response document found linked from the notice.
Hooker, Translational Neurodegeneration
- Main claim
- Reanalysis of CDC data reporting a higher autism risk in African American boys given MMR before 36 months.
- Journal's stated reason
- Undeclared competing interests and concerns about the methods and statistical analysis.
- Author response
- Author response: not linked from the notice.
Inbar, Shoenfeld et al., Vaccine
- Main claim
- Female mice given aluminium adjuvant or Gardasil showed behavioural abnormalities.
- Journal's stated reason
- Withdrawn at the editor-in-chief's request; outside experts judged the methods seriously flawed. A revised version was later published in Immunologic Research.
- Author response
- Authors republished a revised version elsewhere (see notice link).
Li, Tomljenovic, Li, Shaw, J Inorg Biochem
- Main claim
- Aluminium injected at vaccine-adjuvant levels changed immune-gene activity in mouse brains, which the authors compared with autism biomarkers.
- Journal's stated reason
- Retracted at the request of the editor and the authors: gel images in several figures were 'incorrectly presented' and original gels were unavailable, so results were 'clearly not reliable'.
- Author response
- The authors joined the retraction request.
DeLong, J Toxicol Environ Health A
- Main claim
- US women aged 25-29 who had the HPV vaccine were 'less likely to have ever been pregnant' than those who did not.
- Journal's stated reason
- Statement of retraction: post-publication reviews found serious flaws in the statistical analysis and its interpretation.
- Author response
- Author response: not linked from the notice.
Walach et al., Vaccines (MDPI)
- Main claim
- Using Dutch adverse-event reports: 'For three deaths prevented by vaccination we have to accept two inflicted by vaccination.'
- Journal's stated reason
- Misinterpreted adverse-event reports as caused by the vaccine, producing 'incorrect and distorted conclusions'.
- Author response
- Author response: not linked from the notice.
Lyons-Weiler and Thomas, IJERPH
- Main claim
- In one paediatric practice, concluded that unvaccinated children 'are not unhealthier than the vaccinated' and may be healthier overall, based on office visits for many diagnoses.
- Journal's stated reason
- Investigation found methodological issues; conclusions 'not supported by strong scientific data'.
- Author response
- The notice records that the authors did not agree with the retraction.
Skidmore, BMC Infectious Diseases
- Main claim
- From an online survey of 2,840 people, estimated that US deaths from COVID-19 vaccination 'may be as high as 278,000'.
- Journal's stated reason
- Methods could not establish that the vaccine caused the deaths; reported deaths were not verified; limitations not described; ethics statement incorrect.
- Author response
- Skidmore disagreed and published a response. author response
Hulscher, McCullough et al., Forensic Science International
- Main claim
- Systematic review of published autopsy reports, concluding that most of the deaths reviewed were caused by, or significantly contributed to by, COVID-19 vaccination.
- Journal's stated reason
- Withdrawn by the editors-in-chief: inappropriate citation, inappropriate method design, errors and lack of support for the conclusions, and failure to cite disconfirming evidence. An earlier version was removed by Preprints with The Lancet in 2023 'because the study's conclusions are not supported by the study methodology'.
- Author response
- Author response: not linked from the notice.
Mead, Seneff, Wolfinger, Rose, Denhaerynck, Kirsch, McCullough, Cureus
- Main claim
- Review urging 'a global moratorium on the modified mRNA products' because of serious adverse events and an 'unacceptably high harm-to-reward ratio'.
- Journal's stated reason
- Conclusions unreliable: cited references were questioned, and cited references and available data were misrepresented.
- Author response
- The notice records that the authors disagree with the retraction.
Miller, Toxicology Reports
- Main claim
- Of infant deaths reported to VAERS, 58% were within 3 days and 78.3% within 7 days of vaccination.
- Journal's stated reason
- 'Serious methodological flaws' in using VAERS to infer a link, given 'the expected temporal clustering of events independent of causality'.
- Author response
- The author disputes the removal and published the correspondence. author response
Mostert et al., BMJ Public Health
- Main claim
- Excess deaths in 47 Western countries 2020-2022; the discussion named possible causes including COVID-19 vaccination.
- Journal's stated reason
- 'Misinformation in the discussion regarding the possible causes of excess mortality', and limits of the work not sufficiently described.
- Author response
- Two authors agreed with the retraction but not that the article contains misinformation; Mostert disagrees with the decision (stated in the notice).
Removed without a formal retraction notice
- Mawson et al., Frontiers in Public Health 2016 / Journal of Translational Science 2017: Survey of homeschooled children reporting more chronic illness in vaccinated children. Frontiers removed the provisionally accepted paper in 2016. The 2017 journal told Retraction Watch it had retracted it, but the paper is back on the publisher's site and no formal retraction notice was found. Retraction Watch paper
For balance: retractions on the other side
This guide looked for retracted papers that had concluded vaccines are safe or effective. No retraction of a paper with a vaccine-safety conclusion was verified. The nearest case found:
- Ramsay et al., BMC Medicine 2017: Influenza vaccine-effectiveness meta-analysis (repeated vaccination); retracted by its own authors on 12 Aug 2018 for 'erroneous inclusions and an omission of estimates' that affected some conclusions. source
GAP: other retractions sometimes cited (for example papers by Classen, and other Shaw/Tomljenovic papers) were not verified against a notice and are not listed.