Why The Science Has Never Been Settled

The most dangerous phrase in medicine isn’t “we don’t know.” It’s “the science is settled.” History has a lot to say about that.

There is a phrase that gets deployed whenever a conversation starts heading somewhere inconvenient. You will hear it from your doctor, from health journalists, from the kind of person who shares government dietary guidelines with the energy of someone who has personally validated every study cited. The phrase is this: the science is settled.

It is not an invitation to think. It is an instruction to stop.

The implication is that somewhere, in some gleaming institution staffed by serious people in serious coats, a final determination has been made, the argument is closed, and any further questioning is either ignorance or bad faith. You can comply or you can be difficult. Those are the options on offer.

What is less frequently discussed is what this phrase has a habit of preceding. History has a mordant sense of humour about human certainty, and nowhere does it express that humour more consistently than in the field of medicine. The story of what we once knew for certain, and what that certainty cost the people subjected to it, is not a story of fringe quackery. It is the story of mainstream consensus, institutional authority, and confident experts doing catastrophic things to trusting patients before eventually, reluctantly, and often decades too late, admitting they had it wrong.

What follows is a brief tour of that history. Not to suggest that medicine is useless, because it is not, but to suggest that “settled” is a word that the evidence does not support, has never supported, and that the people harmed along the way might have had something to say about.

The Man Who Washed His Hands and Paid for It

In 1847, a Hungarian physician named Ignaz Semmelweis was working in the maternity ward of the Vienna General Hospital and noticing something that troubled him. Women giving birth in the ward staffed by medical students and physicians were dying of puerperal fever at a rate dramatically higher than those in the ward staffed by midwives. The difference was consistent. It was measurable. It was, to Semmelweis, clearly pointing at something.

What he eventually identified was this: the medical students and physicians were coming directly from performing autopsies and going straight to delivering babies without washing their hands. The midwives were not performing autopsies. Semmelweis began requiring that his staff wash their hands in a chlorinated lime solution before examinations. The mortality rate in his ward dropped from around ten per cent to under two per cent almost immediately.

The medical establishment’s response was to reject him.

His colleagues found the idea offensive. The implication that physician hands could be carrying something lethal, that doctors were contributing to their patients’ deaths simply by touching them, was incompatible with the prevailing theory that disease arose from miasmas, from bad air, from constitutional weakness. Some found it an affront to professional dignity. Semmelweis was not temperate in his frustration, which did not help his cause, and the criticism intensified until he suffered what appears to have been a mental breakdown in 1865. He was committed to a psychiatric institution. He died two weeks later, likely from the very kind of infection he had spent his career trying to prevent.

Germ theory, the framework that would have explained and vindicated everything he observed, was formalised by Pasteur and Koch in the years immediately following his death. Within a decade, the medical community had accepted the existence of pathogens. Handwashing became standard. Semmelweis was posthumously rehabilitated into a hero.

He died in an asylum. The science, as it turned out, had not been settled. It had merely been defended, forcefully and at considerable human cost, by people with too much institutional prestige invested in being right.

Two Thousand Years of Confident Bloodletting

Before we give the nineteenth century too much of a hard time, it is worth acknowledging that the most durable medical consensus in recorded history lasted roughly two thousand years and involved draining blood from sick people.

The theory, derived from Galen and the ancient Greeks, held that the body contained four humours: blood, phlegm, black bile, and yellow bile. Disease was the result of these falling out of balance. The treatment was to correct the imbalance, and for a great many conditions, that meant removing blood. Physicians carried lancets. They applied leeches. They bled patients who were already weak, already feverish, already physiologically compromised, in the confident belief that they were helping.

George Washington, the first President of the United States, fell ill with a throat infection in December 1799. His physicians, following best practice, removed roughly two and a half litres of blood over the course of a day, approximately half his total blood volume, before he died. They did not consider themselves negligent. They were applying the standard of care. The science, by any measure of that era, was settled.

Bloodletting persisted in mainstream medicine well into the nineteenth century. It was not some fringe practice of hedge physicians and travelling charlatans. It was taught in universities, conducted in hospitals, and defended by the most eminent practitioners of the day. When the evidence against it began to accumulate, it retreated slowly, not because the establishment was persuaded by the data but because a generation of practitioners who had built careers on it gradually died off and were replaced by people who had not.

Two thousand years. Bear that in mind the next time someone tells you something has been “well established.”

The Nobel Prize for the Ice Pick

In 1949, the Nobel Committee awarded the Prize in Physiology or Medicine to Antonio Egas Moniz for the development of the prefrontal leucotomy. You will know it by its more familiar name: the lobotomy.

The procedure involved severing the connections between the prefrontal cortex and the rest of the brain, initially through holes drilled in the skull, and later, in the hands of the American psychiatrist Walter Freeman, via an ice pick inserted through the eye socket and moved back and forth. Freeman was so enthusiastic about the technique that he performed it on an outpatient basis, toured the country in a van he called the lobotomobile, and conducted the procedure in hotel rooms.

He performed over three thousand lobotomies. One of his patients was Rosemary Kennedy, John F. Kennedy’s sister, who went in with mood instability and came out permanently incapacitated, requiring institutional care for the rest of her life.

At the time, this was medicine. Psychiatry had no effective pharmacological treatments for severe mental illness. Lobotomy produced patients who were calmer, more manageable, less likely to disturb the institutional routine. That it also frequently produced patients who were incapable of complex thought, had lost their personalities, or were simply warehoused in a vegetative state was considered an acceptable outcome, or was not considered at all.

The Nobel Prize was never rescinded.

Psychopharmacology, when it arrived in the 1950s, did not make lobotomy obsolete because the scientific establishment reviewed the evidence and updated its position. It made lobotomy obsolete because there was now a more convenient alternative. The lesson of the lobotomy era is not simply that one bad treatment was replaced by better ones. It is that an intervention which caused profound and irreversible harm to tens of thousands of people was endorsed at the highest level of scientific recognition while it was happening. The committee did not give the prize to a maverick. They gave it to the consensus.

The Morning Sickness Drug That Stopped Limbs Growing

In the 1950s, a West German pharmaceutical company developed a drug that proved remarkably effective for morning sickness. Thalidomide was marketed across Europe, Canada, and Australia as safe and well-tolerated. It was prescribed to pregnant women. It crossed the placental barrier. In the children born to mothers who had taken it, it disrupted limb development. Around ten thousand children were born with phocomelia: severely shortened or absent limbs. Thousands more died before birth.

The drug had not been tested on pregnant animals before it was released. The assumption was that safety in adults translated to safety in pregnancy. The assumption was wrong, catastrophically and irreversibly so for every family affected.

The United States was largely spared because a single FDA reviewer named Frances Kelsey refused to approve the drug without adequate safety data, holding out against considerable commercial pressure. Her obstruction was vindicated. She received a presidential award for distinguished federal civilian service.

In the countries where thalidomide was approved, it had passed regulatory review. It had the endorsement of the relevant authorities. The manufacturers were confident in its safety profile. None of that confidence made a single arm grow back.

The Doctor Will See You Now. Have a Cigarette.

For much of the first half of the twentieth century, tobacco companies faced a public relations problem. People were beginning to suspect, with mounting evidence behind them, that smoking might not be entirely good for the lungs. The industry’s response was a masterclass in manufactured consensus.

Physicians were recruited to appear in advertising campaigns. “More doctors smoke Camels than any other cigarette,” ran one of the most famous, accompanied by a reassuring man in a white coat. Chesterfield ran advertisements claiming their cigarettes were “just as pure as the water you drink.” The American Tobacco Company marketed Lucky Strikes as a weight management aid, encouraging women to reach for a cigarette instead of a sweet. Physicians were photographed, quoted, and paid. Medical associations accepted tobacco industry funding. Journals ran advertisements from cigarette brands alongside clinical research.

When the evidence of harm became impossible to ignore, the industry did not concede. It produced its own research, seeded doubt about the methodology of the studies linking smoking to cancer, and coined the strategy of manufacturing uncertainty that would later be recycled, with minor modifications, by the fossil fuel industry, the sugar industry, and several pharmaceutical companies. The playbook was so effective that it has been studied as a template ever since.

The scientific consensus that smoking was safe, or at worst neutral, was not fringe opinion. It was actively constructed, expensively maintained, and endorsed by people whose professional authority should have made it unassailable. By the time the US Surgeon General’s report finally landed in 1964, the evidence had been available, suppressed, and argued over for the better part of two decades. Lung cancer rates had been climbing throughout. The doctors in the advertisements did not cause all of that harm personally. But they were the face of a consensus that cost millions of lives, and they were paid to be.

The Drug That Rewired an Opioid Crisis

In 1996, Purdue Pharma launched OxyContin with a marketing campaign built on a single claim: it was less addictive than conventional opioids because its extended-release formulation meant a smooth, steady delivery rather than the peaks and troughs that drove dependency. This claim was made to physicians across the United States with the confidence of a company that had paid for the studies that supported it.

What followed was the opioid epidemic. Physicians prescribed OxyContin at extraordinary rates, reassured by the safety narrative. Patients became dependent. The pills were crushed and snorted or dissolved and injected when swallowed whole no longer delivered the required effect. People who had never sought anything stronger than aspirin found themselves physically dependent on opioids within weeks. The transition to street heroin, when OxyContin became expensive or unavailable, was well documented and numerically vast.

By the time the full scale of the disaster was clear, hundreds of thousands of Americans had died of opioid overdoses. Purdue Pharma eventually pleaded guilty to criminal charges. The Sackler family, who owned the company, paid billions in settlements while denying personal wrongdoing.

Every prescription that started this chain was written by a physician acting on medical consensus. The science, at the time of prescribing, appeared settled. Addiction specialists who raised concerns early were characterised as catastrophising. Pain was undertreated, they were told. The new drugs were safe. The data said so. The company said so. The authorities who approved it said so.

It is worth sitting with this example a little longer than the historical ones, because the gap between what was officially asserted and what was demonstrably true was not separated by decades of subsequent research. It was visible, to anyone willing to look, almost immediately. The question is not simply whether the science was settled. The question is who settled it, and whose interests were served by it remaining settled.

The Dietary Fat Catastrophe: A Story You’ve Heard Before, And Should Hear Again

There is a longer article on this site that covers the Seven Countries Study and the career of Ancel Keys in the detail it deserves. The short version, for the purposes of this particular argument, is this.

In the middle of the twentieth century, a hypothesis emerged that dietary fat, and saturated fat in particular, was the primary driver of cardiovascular disease. This hypothesis was contested at the time by other researchers, who pointed out the methodological weaknesses in the supporting evidence. Those researchers lost the argument, not because the evidence was resolved in Keys’s favour, but because Keys was more politically adept, better connected, and better positioned to influence the committees and institutions that converted hypothesis into policy.

The result was a set of dietary guidelines, first issued in the United States in 1980 and replicated across the Western world, instructing populations to reduce fat intake and replace it with carbohydrates. The food industry, responding to this guidance, reformulated thousands of products to remove fat and maintain palatability through sugar. Seed oils, extracted from plants using industrial processes that would have been unrecognisable to any previous generation of humans, flooded the food supply as the supposedly heart-healthy replacement for animal fats.

Obesity rates, which had been relatively stable, began rising sharply in the years following the introduction of the guidelines. Type 2 diabetes followed. Metabolic syndrome became a primary concern of public health. The diseases the guidelines were designed to prevent accelerated in almost perfect correlation with compliance.

This is not a retrospective fringe position. The evidence against the dietary fat hypothesis has accumulated to the point where the original researchers who contested it, dismissed for decades as cranks and contrarians, have been substantially vindicated. The sugar industry’s funding of research that redirected blame from sugar to fat has been documented in the historical record. The financial relationships between food manufacturers and the institutions that produced the guidelines have been traced in peer-reviewed papers.

The guidelines remain largely in place. The consensus moves slowly when there are industries whose revenue depends on its not moving at all.

Which Brings Us, Delicately, To Statins

Statins are the most prescribed class of drugs in the world. They lower LDL cholesterol. The prevailing consensus holds that lowered LDL cholesterol reduces cardiovascular risk, and that therefore statins save lives. This is presented as settled.

It is worth applying the same forensic attention to this consensus that history would suggest is appropriate.

The clinical trials that established statin efficacy were funded, in large part, by the pharmaceutical companies that manufacture statins. Independent reanalyses of the underlying trial data have been complicated by the fact that the raw data has not been made available to independent researchers, which is an unusual position for a scientific conclusion claimed to be beyond dispute. The trials that showed benefit were published. The trials that did not show benefit have a documented history of non-publication.

The metric used to sell statins to patients is relative risk reduction, which produces the most impressive-sounding numbers. The metric more relevant to an individual patient is absolute risk reduction, or Number Needed to Treat: how many people need to take this drug for one person to avoid the outcome it is designed to prevent. For primary prevention in low-risk patients, that number is in the range of one hundred to two hundred. Which means that for every patient the statin helps, ninety-nine to one hundred and ninety-nine are taking it daily, indefinitely, experiencing its side effects, for no measurable personal benefit.

The side effects are not trivial. Muscle pain, classified as myalgia, affects a meaningful percentage of patients. Statins have been associated with increased risk of type 2 diabetes. Cognitive effects have been reported sufficiently often to result in FDA label changes. These are not invented by contrarians. They are in the drug’s own documentation.

None of this means statins are useless. For secondary prevention in high-risk patients, the evidence of benefit is considerably stronger. But “the science is settled” is not a position that the full evidence record supports, and the conflicts of interest embedded in that evidence record are not categorically different from the conflicts of interest that gave us the dietary fat hypothesis, the opioid epidemic, and every other consensus that later turned out to have been somewhat more complicated than advertised.

The appropriate response to statin prescribing is not blanket refusal. It is the same response that would have been appropriate to bloodletting, to lobotomies, to thalidomide, to OxyContin at each stage of their histories: a willingness to ask questions, examine the evidence independently, and decline to mistake institutional authority for proof.

Why It Keeps Happening

The pattern across all of these examples is not, primarily, one of malice. The bloodletters were not sadists. The lobotomists believed they were helping. The physicians who prescribed OxyContin were not in most cases corrupt. The pattern is structural, and it repeats because the structure never changes.

Medical careers are built on consensus. Challenging it is professionally expensive. The researcher who questions the dominant paradigm does so knowing that funding, publication, and reputation all flow more easily to work that confirms existing frameworks than to work that undermines them. Semmelweis did not lose his position because he was wrong. He lost it because being right required everyone around him to accept that they had been killing patients. Human institutions are not designed to absorb that kind of correction gracefully.

Add to this the funding architecture of modern medicine: the clinical trials that establish drug safety and efficacy are overwhelmingly funded by the companies that profit from positive results. The researchers who conduct them frequently have financial relationships with those companies. The journals that publish them depend on advertising revenue from the same industry. The regulators who approve the resulting drugs are often drawn from, and return to, the industries they regulate. None of this requires individual bad faith. All of it produces systematic bias.

And then there is the problem of time. The tobacco industry manufactured doubt about smoking for two decades while lung cancer rates climbed. Thalidomide was withdrawn in 1961. The dietary fat consensus was embedded in policy in 1980. The opioid epidemic began in earnest in the 1990s. In each case, the harm was measurable within years. The consensus did not move within years.

It moved on timescales measured in decades, and it moved only when the accumulated weight of the contrary evidence became too large to contain, or when the legal consequences of not moving became too significant to absorb. The people harmed in the interval between harm being demonstrable and consensus being revised were not a rounding error. They were the entire point.

The Only Sensible Response

None of this is an argument for rejecting medicine wholesale, for treating every doctor as a functionary of some malevolent conspiracy, or for assuming that whatever the mainstream recommends is automatically wrong. Medicine has produced genuine miracles. Germ theory, vaccination, surgical technique, anaesthesia, antibiotics: the list of things that work, that have demonstrably extended and improved human life, is long and real.

The argument is narrower and more specific. It is that “settled science” is a phrase with a history, and that history is not a comfortable one. Every consensus that later turned out to be wrong was, at the time it was wrong, presented as settled. The people who questioned it were, at the time they were questioning it, dismissed as dangerous. The harm that accumulated while the consensus held was, at the time it was accumulating, explained away or not examined at all.

The appropriate response to any health claim, any dietary guideline, any pharmaceutical recommendation, is not deference and not rejection. It is the same thing it has always been: ask who funded the research, ask what the absolute numbers say rather than the relative ones, ask which results were published and which were not, and ask what the people who disagree with the consensus actually argue, rather than how the consensus characterises their argument.

The science is never settled. It is always in process, always partial, always subject to revision by evidence that has not yet been gathered or has not yet been heard. The scientists who said otherwise, throughout history, were the ones who ended up on the wrong side of it.

The next time someone tells you the evidence is in, it might be worth asking: in for whom, funded by whom, and settled at what cost to the people who trusted that it was.

The Settled Science Series

This is part of an ongoing series on the need for healthy skepticism when looking at mainstream science.

• How Seed Oils Poisoned The Planet
• The Corrupt History Behind The Dietary Guidelines
• Why Nutritional Science Is A Waste Of Time
• The Industrial Origins Of Skin Cancer

If you want to put together your own metabolic revival, follow me on Twitter or Instagram, or find out what the coaching programme involves below.