Medical Reversals: 130 Years of Health Advice We Later Regretted

Medical reversals are one of the healthiest reminders that medicine is practiced by human beings, not by an infallible machine. Over the last 130 years, treatments, products, and public-health claims have sometimes moved from celebrated innovation to cautionary tale. The lesson is not that science is useless. The lesson is that evidence matures, incentives matter, long-term harms can hide behind short-term benefits, and confidence should never outrun what the data can actually prove.

That distinction matters because not every historical mistake was the same. Some were pharmaceutical marketing failures. Some were regulatory failures. Some were widely accepted medical practices built on weak evidence. Some were plausible ideas that failed once tested against outcomes that mattered. Good science eventually corrected many of them — but patients often paid the price before the correction arrived.

For Six Stone Jars, the practical posture is neither blind trust nor reflexive distrust. It is health stewardship: ask what the intervention does, what outcome was actually measured, how large the absolute benefit is, what harms are known, what remains unknown, who funded the evidence, and whether the recommendation fits your personal risk. That same framework underlies What Does It Mean to Be Healthy?, where health is treated as whole-person stewardship rather than passive compliance.

The Viral List Gets Something Important Right — but Not Everything

A recent social-media post compressed 130 years of medical history into a provocative timeline. Its central warning is legitimate: qualified people and respected institutions have sometimes been confidently wrong. But several lines mix accurate history with oversimplification. That is exactly why this subject deserves a slower, evidence-based review.

Claim from the Viral List	Verdict	What the Evidence Supports
Bayer marketed heroin as a nonaddictive morphine substitute and cough medicine, including for children.	Substantially true	Heroin was marketed from the late 1890s as a safe/nonaddictive morphine substitute and antitussive; historical literature documents pediatric cough use.
Calomel teething powders caused pink disease in infants.	True	Mercury-containing calomel powders were a major cause of infantile acrodynia; the disease largely disappeared after these products were withdrawn.
Crisco’s hydrogenated cottonseed oil introduced a fat later banned as unsafe.	Core point true, wording needs care	Crisco launched in 1911 using hydrogenated cottonseed oil. Artificial trans fat from partially hydrogenated oils was later found to increase cardiovascular risk; FDA removed PHOs from GRAS status.
Radithor radioactive water was promoted for vitality and caused catastrophic injury.	True	Radium tonics were marketed as health products; Eben Byers’ well-documented radiation poisoning became a major public-health warning.
Elixir Sulfanilamide dissolved in diethylene glycol killed 107 people and helped create modern drug-safety law.	True	FDA records 107 deaths, many children, and links the disaster directly to the 1938 law requiring premarket drug-safety evidence.
Doctors and medical journals promoted cigarettes.	True	Doctor-targeted cigarette advertising appeared in major medical journals, including JAMA; physician imagery and ‘more doctors smoke Camels’ messaging were used to reassure consumers.
DDT was broadly used before environmental and toxicological harms drove restriction.	True with nuance	DDT had real public-health benefits against vector-borne disease, but widespread agricultural/domestic use produced persistence, biomagnification, ecological harm, resistance, and toxicological concern.
Thalidomide for pregnancy nausea caused more than 10,000 severe birth-defect cases.	True	The catastrophe involved more than 10,000 children worldwide and transformed drug regulation.
The 1970s margarine switch proves medical advice caused trans-fat harm.	Partly true / oversimplified	Partially hydrogenated fats were widely consumed and later restricted, but ‘margarine’ was not one uniform product and the history cannot be reduced to a single AHA instruction.
The 1980s low-fat era caused obesity because sugar replaced fat.	Plausible contribution, not proven as a single cause	Some low-fat processed foods increased refined carbohydrate/sugar, but the obesity epidemic is multifactorial and cannot be attributed to one dietary recommendation.
OxyContin was marketed as less addictive and Purdue later pleaded guilty to misbranding.	True	Federal cases document false or misleading marketing about addiction, abuse, and dependence risk. The famous ‘less than 1%’ claim has a more complicated lineage than the viral post implies.
Vioxx was withdrawn after cardiovascular risk emerged; an FDA safety officer estimated tens of thousands of excess events.	True	Vioxx was withdrawn in 2004; FDA epidemiologist David Graham testified to an estimated range of 88,000–139,000 excess serious coronary events under one set of assumptions.

What These Reversals Have in Common

The examples are different, but several recurring failure modes appear again and again.

• A plausible mechanism is mistaken for proof of long-term benefit.
• A surrogate marker improves, but the outcome patients care about gets worse.
• Short trials miss harms that take years, pregnancy, aging, or cumulative exposure to reveal.
• Commercial incentives amplify confidence faster than evidence justifies.
• Regulators and professional groups inherit the same assumptions as the industry they oversee.
• Patients are told relative benefit without being shown absolute benefit.
• Rare harms are invisible until millions of people are exposed.
• Once a practice becomes standard, questioning it can feel socially or professionally disloyal.

One of the cleanest demonstrations came from the Cardiac Arrhythmia Suppression Trial (CAST). Encainide and flecainide successfully suppressed ventricular premature beats after heart attack — exactly what clinicians hoped they would do — yet patients receiving the drugs had substantially higher arrhythmic death and total mortality. The surrogate improved. The patients did worse.

“The most dangerous medical mistake is not being wrong. It is being more certain than the evidence allows.”

A Closer Look at Today’s Medical Watch List

History naturally raises the question: which current practices will future generations criticize? No one can know in advance. If we could, they would already be reversals. But we can identify areas where evidence, dose, long-term exposure, absolute benefit, or informed consent deserve unusually careful attention.

1. Community Water Fluoridation: Benefit, Dose, and Neurodevelopment

The strongest current evidence does not support the statement that fluoride ‘does not prevent cavities.’ A 2024 Cochrane review concluded that contemporary community water fluoridation may produce a small reduction in tooth decay in children, although the effect appears smaller than in pre-1975 studies and evidence in adults was lacking.

The neurodevelopment question is more serious than older public messaging often acknowledged. The National Toxicology Program’s 2024 monograph concluded with moderate confidence that higher fluoride exposure — particularly around or above 1.5 mg/L — is associated with lower IQ in children. But NTP also states there are insufficient data to determine whether the 0.7 mg/L level recommended for U.S. community water has that effect, and it found no evidence of adverse adult cognition.

That makes fluoride a good example of how the right question is often dose-specific rather than ideological. The evidence justifies continued research and transparent discussion; it does not currently prove that normal U.S. fluoridation causes dementia or neurological decline in old age.

2. Dental Amalgam: Mercury Exposure Without an All-or-Nothing Conclusion

Dental amalgam is approximately 50% elemental mercury by weight and releases small amounts of mercury vapor. The FDA recommends avoiding new amalgam when possible in certain higher-risk groups, including pregnant women, young children, people with kidney disease, neurological disease, or mercury sensitivity.

At the same time, FDA states that available evidence does not show that amalgam exposure causes adverse health effects in the general population and specifically advises against removing intact, functional amalgam solely to prevent disease, because removal temporarily increases mercury exposure and sacrifices healthy tooth structure.

That is more nuanced than saying ‘mercury fillings are proven toxic to everyone.’ The more defensible concern is that a mercury-containing material is still used despite identifiable susceptible groups and limited long-term data in some populations — making informed choice and alternatives important.

3. Vaccines and Mercury: The Ingredient Claim Needs Precision

The statement that routine vaccines are broadly ‘filled with mercury’ is no longer accurate in the United States. CDC states that most vaccines contain no mercury. Thimerosal, an ethylmercury-containing preservative, was removed from or reduced in routine childhood vaccines beginning in 2001; it remains in some multi-dose influenza formulations and a limited number of other products.

CDC also reports that studies have not found a link between thimerosal-containing vaccines and autism or neuropsychological delay. Ethylmercury is not the same exposure as methylmercury from contaminated food.

None of that means vaccine risk should be treated casually. Vaccines are medical products and can have adverse effects, contraindications, age-specific risk, and product-specific differences. The correct standard is transparent benefit-risk information for the actual vaccine, not blanket claims that all vaccines are either perfectly harmless or broadly toxic.

4. COVID-19 Vaccines: A Real Example of Risk Communication That Changed

COVID-19 vaccine communication is a more useful case study when stated precisely. The claim that the vaccines ‘never prevented infection or transmission’ is contradicted by early randomized and real-world evidence. The original Pfizer trial reported about 95% efficacy against symptomatic COVID-19, and large observational studies found strong early protection against documented infection. Household studies also found reduced transmission from vaccinated individuals.

But those benefits were not permanent. Protection against infection waned with time and changed substantially as variants evolved. Meanwhile, post-market surveillance identified rare safety signals that deserved explicit risk communication.

One is myocarditis and pericarditis after mRNA vaccination, which CDC recognizes as causally associated and most frequent in adolescent and young adult males, especially shortly after a second dose. Later guidance became more age-, sex-, interval-, and risk-specific.

The historical lesson is not ‘the vaccines were neither safe nor effective.’ That goes beyond the evidence. The more defensible criticism is that public messaging sometimes sounded more absolute than the evolving data justified, especially around transmission, waning immunity, product-specific risk, and the need for individualized benefit-risk decisions.

5. Statins for Primary Prevention: Absolute Benefit Matters

Statins are highly effective for many people with established cardiovascular disease and can benefit higher-risk people who have not yet had an event. The debate becomes more important in low- or intermediate-risk primary prevention, where relative risk reductions may sound larger than the absolute benefit.

The U.S. Preventive Services Task Force evidence review found an absolute reduction of about 0.85–0.89 percentage points in fatal or nonfatal myocardial infarction across pooled primary-prevention trials, with larger absolute benefit in people at higher baseline risk.

So ‘one heart attack prevented per hundred people’ is not a universal statin fact, but it is in the neighborhood of some pooled primary-prevention estimates over several years. The right question is not whether statins are good or bad. It is: What is this person’s baseline risk, what absolute benefit is expected, what dose is needed, what adverse effects matter, and are lifestyle measures being addressed at the same time?

6. GLP-1 and GIP/GLP-1 Weight-Loss Drugs: Powerful Benefit, Real Lean-Mass Questions

Modern incretin drugs can produce major weight loss and important metabolic benefits. They also raise reasonable questions about long-term dependence, gastrointestinal effects, gallbladder disease, cost, rebound after discontinuation, nutrition, and preservation of lean tissue.

A 2025 body-composition substudy of tirzepatide found that about 25% of weight lost was lean mass and about 75% was fat mass. Lean mass is not identical to skeletal muscle, and some lean-mass loss accompanies ordinary weight loss as well, so the viral claim that ‘a third of the loss is muscle’ is too blunt.

Still, preserving strength and muscle is especially important for older adults and people at risk of frailty. That makes protein intake, resistance training, dose selection, and long-term maintenance worthy of more attention as these drugs move from specialist care into mass use.

7. Polypharmacy, Sedatives, and the Cascade Problem

One of the strongest candidates for future criticism may be less dramatic than a new technology: the accumulation of prescriptions. Older adults can end up taking one drug for a disease, another for the first drug’s side effect, another for sleep, another for anxiety, and supplements on top.

The problem is not that each medicine is necessarily inappropriate. It is that evidence for an individual drug does not automatically establish safety for a ten-drug combination in an 80-year-old patient with changing kidney function, balance, cognition, and nutritional status.

Long-term benzodiazepine and sedative use deserves particular caution because dependence, falls, cognitive effects, and withdrawal can be serious. Regular medication review and deprescribing when appropriate may eventually be seen as one of the most important corrections in geriatric care.

8. Antibiotic Overuse and the Microbiome

Antibiotics transformed medicine and remain lifesaving. Their success also made overuse easy. Unnecessary antibiotic exposure selects for resistant organisms, can cause direct adverse effects, and alters the microbiome in ways science is still learning to interpret.

This is another case where the future correction is unlikely to be ‘antibiotics were bad.’ It will be about precision: correct indication, narrowest useful drug, shortest effective course, and greater respect for ecological effects inside and outside the body.

9. Screening, Incidental Findings, and Overdiagnosis

Modern imaging and screening can find disease earlier than ever. That is often valuable. It can also detect abnormalities that would never have caused symptoms during a person’s lifetime.

Overdiagnosis can lead to biopsy, surgery, radiation, anxiety, lifelong labeling, and treatment complications without improving survival. Thyroid cancer, prostate cancer, pulmonary nodules, and other incidental findings have all forced medicine to confront the difference between detecting more disease and helping more patients.

This may become one of the defining medical reversals of the precision-medicine era: recognizing that seeing more is not always the same as knowing what should be treated.

What About Seed Oils, Hexane, and Infant Formula?

The viral post ends by asking whether future generations will be shocked by hexane-extracted seed oils and infant formula made with vegetable oils. These are legitimate topics for research, but they do not currently belong in the same evidentiary category as thalidomide, Elixir Sulfanilamide, or Vioxx.

Industrial extraction methods, oxidation products, fatty-acid balance, and ultra-processed food exposure are reasonable subjects for nutrition research. But evidence that normal residual hexane exposure from edible oils is causing widespread human disease is not established. Likewise, infant formula commonly uses blends of vegetable oils to supply required fats and essential fatty acids; formula is a regulated nutritional product and can be medically essential. That does not place every formulation above scrutiny, but ‘contains vegetable oil’ is not evidence of harm.

This is an important discipline for skepticism: do not promote a future reversal before the reversal exists.

“Good skepticism asks harder questions. Bad skepticism announces the answer before the evidence arrives.”

A Better Way to Evaluate Medical Advice

Question	Why It Matters	Practical Use
What outcome was measured?	A lab marker or surrogate may not predict survival, function, pain, or quality of life.	Prefer evidence on outcomes patients actually care about.
What is the absolute benefit?	Large relative percentages can describe small real-world differences.	Ask for absolute risk reduction and number needed to treat.
What are the important harms?	Rare harms become visible only after large-scale use.	Ask about frequency, severity, reversibility, and who is at highest risk.
How long were patients studied?	Short trials can miss cancer, neurodevelopmental, metabolic, fertility, and cumulative effects.	Match follow-up length to the kind of harm that could plausibly occur.
Who was studied?	Trial populations may not match older adults, children, pregnant women, or people with multiple conditions.	Ask how closely the evidence fits you.
Who funded and designed the study?	Funding does not invalidate evidence, but conflicts can influence design, endpoints, publication, and messaging.	Look for replication and independent evidence.
Are alternatives being compared fairly?	A treatment can beat placebo yet still be inferior to lifestyle, watchful waiting, or a safer therapy for some patients.	Ask what happens with no treatment and with reasonable alternatives.
Can the decision be revisited?	Some choices are reversible; others are not.	Prefer staged decisions when uncertainty is high and urgency is low.

The Lesson Is Not ‘Never Trust Science’

It is tempting to look at this history and conclude that experts cannot be trusted. That would be the wrong lesson. The same scientific method that allowed errors to persist also exposed them. Randomized trials overturned antiarrhythmic practice. Epidemiology exposed smoking and Vioxx risk. Toxicology and surveillance identified mercury poisoning. Pharmacovigilance transformed drug regulation after thalidomide.

The better lesson is that trust should be earned continuously. Science should be open to challenge. Regulators should disclose uncertainty. Companies should not be allowed to turn relative risk into marketing theater. Professional societies should distinguish evidence from consensus. Patients should be treated as adults capable of understanding both benefit and harm.

That is also why the approach in How to Prevent Commercial Interests from Stealing Your Health is useful: commercial incentives are not proof of wrongdoing, but they are a reason to keep informed consent, transparency, and independent verification strong.

The Bottom Line

The last 130 years contain enough genuine medical reversals that we do not need to exaggerate them. Heroin cough syrup, calomel teething powders, radioactive tonics, toxic drug solvents, doctor-endorsed cigarette advertising, thalidomide, lethal antiarrhythmic strategies, deceptive opioid marketing, and Vioxx are already powerful reminders of medical fallibility.

Today’s lesson is not to assume that every current intervention will become tomorrow’s scandal. It is to recognize the conditions that make reversals more likely: weak long-term evidence, small absolute benefit, high commercial pressure, surrogate endpoints, vulnerable populations, cumulative exposure, poor adverse-event surveillance, and a culture that punishes reasonable questions.

Good medicine needs good science. Good science needs humility. And good health stewardship requires patients who are informed enough to ask for both.

Closing CTA

Before starting your next long-term medication, procedure, supplement, or preventive intervention, ask five questions: What is my absolute expected benefit? What are the known harms? What remains uncertain? What alternatives exist? When will we reassess whether this is still helping? Better questions do not undermine medicine. They make medicine safer.

Source and Fact-Checking Notes

• FDA — Sulfanilamide Disaster — Confirms more than 100 deaths from Elixir Sulfanilamide containing diethylene glycol and the role of the disaster in the 1938 drug-safety law. Source
• FDA — Milestones in U.S. Food and Drug Law — Documents the 107 Elixir Sulfanilamide deaths, 1938 safety law, thalidomide regulatory history, and later drug-safety milestones. Source
• PubMed / CMAJ — Heroin History — Documents Bayer-era marketing of heroin as a safe/nonaddicting morphine substitute. Source
• PubMed — The Rise and Fall of Pink Disease — Historical review of infantile acrodynia caused by mercury-containing teething powders. Source
• FDA — Partially Hydrogenated Oils — Confirms 2015 determination that PHOs were no longer GRAS because artificial trans fat raises cardiovascular risk. Source
• EPA — DDT: A Brief History and Status — Documents DDT’s public-health usefulness, widespread use, environmental persistence, toxicological concerns, and U.S. cancellation. Source
• FDA — Frances Kelsey / Thalidomide — Documents the thalidomide catastrophe and regulatory consequences. Source
• PubMed — CAST — Landmark randomized evidence that encainide/flecainide suppressed arrhythmias yet increased arrhythmic death and total mortality. Source
• DOJ — Purdue / OxyContin Misbranding — Documents false marketing that OxyContin was less addictive, less subject to abuse/diversion, and less likely to cause dependence/withdrawal. Source
• FDA Testimony / Graham Vioxx Estimate — Historical testimony estimating tens of thousands of excess serious coronary events associated with Vioxx. Source
• Cochrane — Water Fluoridation 2024 — Contemporary evidence suggests a small caries benefit in children, smaller than pre-1975 estimates; adult evidence lacking. Source
• NTP — Fluoride Neurodevelopment and Cognition — Moderate confidence that higher fluoride exposures are associated with lower child IQ; insufficient data to determine effect at 0.7 mg/L; no evidence of adverse adult cognition. Source
• FDA — Dental Amalgam Fillings — Documents mercury composition, higher-risk groups, general-population evidence, and recommendation not to remove intact fillings solely to prevent disease. Source
• CDC — Thimerosal and Vaccines — States most routine childhood vaccines no longer contain thimerosal and research has not shown an autism or neuropsychological-harm link. Source
• NEJM — Pfizer mRNA Vaccine Trial — Documents strong early efficacy against symptomatic COVID-19. Source
• NEJM — Vaccination and Household Transmission — Documents reduction in household COVID-19 among contacts of vaccinated health-care workers. Source
• CDC — Myocarditis After COVID-19 Vaccines — Recognizes rare causal association, especially in adolescent/young adult males after mRNA vaccination. Source
• USPSTF — Statins for Primary Prevention — Provides pooled relative and absolute benefit estimates and emphasizes baseline-risk dependence. Source
• PubMed — Tirzepatide Body Composition — Reports roughly 75% fat-mass and 25% lean-mass contribution to weight loss in a SURMOUNT-1 substudy. Source

Sources and Further Reading

• FDA — Promoting Safe & Effective Drugs for 100 Years — Read
• EPA — DDT Regulatory History — Read
• NHLBI — Women’s Health Initiative — Read
• CDC — COVID-19 Vaccine Safety — Read
• FDA — Information for Patients About Dental Amalgam — Read
• USPSTF — Statin Recommendation — Read

Disclosure

This article is educational and is not personal medical or dental advice. Historical medical reversals do not prove that current medical recommendations are wrong, and unresolved scientific questions should not be treated as established harm. Decisions about prescription drugs, vaccines, dental restorations, fluoridated water, weight-loss medication, screening, or other medical interventions should be made using individual risk, current evidence, and qualified professional guidance. Do not stop prescribed medication, remove sound dental amalgam, or alter a vaccination plan solely because of this article.