An FDA panel backed unproven peptides over its own scientists

Biotech & Longevity 3 min 80 sources

The week's biggest biotech story was a regulatory one - a US panel voting to widen access to barely-tested peptides. Alongside it: two large "linked to" health studies, a weight-loss drug that fell short on the heart, and a record measles year.

Key takeaways

An FDA advisory panel voted to loosen access to popular but unproven peptides, overruling the agency's own scientists.
Two big studies "linked" coffee to healthier livers and road noise to Parkinson's, but a link is not proof of a cause.
Lilly's retatrutide drove up to 22.6% weight loss yet failed to show it lowers heart risk, and US measles cases hit a three-decade high.

The panel that overruled its own scientists

A US Food and Drug Administration advisory panel voted this week to loosen restrictions on a group of peptides that are popular online but barely tested in people. [34] The panel narrowly backed letting specialised pharmacies make drugs including BPC-157, TB-500 and KPV - a major hurdle cleared toward selling them by prescription. [34][45]

Peptides are short chains of amino acids, small proteins the body makes naturally to carry messages between cells. [34] Some, like insulin, are long-established medicines. But injectable peptides sold for healing, anti-ageing and recovery have spread fast online, marketed by influencers, since GLP-1 weight-loss drugs went mainstream. [34]

The votes, several of them 8-6, went against the FDA’s own scientists, who told the two-day hearing the drugs had not been shown to be safe or effective. [28][35] Several panel members were themselves sellers of experimental peptides. [28] The recommendation is not binding, but the agency often follows its panels. Health Secretary Robert F. Kennedy Jr., who oversees the FDA, has called himself a “big fan” of peptides and vowed to ease the earlier restrictions. [35][28] For a market built largely on influencer testimony, it is a big step toward the pharmacy counter. [50]

What this week’s “linked to” headlines showed

Two large studies landed with familiar-sounding health headlines. Both are worth reading slowly.

Cedars-Sinai researchers followed nearly 355,000 UK adults for a median of 13 years and found coffee drinkers had healthier livers. [26] Compared with non-drinkers, those on five or more cups a day had a 32% lower risk of cirrhosis and a 47% lower risk of liver cancer. [26] Their risk of dying from liver disease was 42% lower. [26] Liver scans and blood proteins pointed the same way, showing less fat, scarring and inflammation. [26]

The second, in JAMA Neurology, tracked 3.1 million Danes aged 40 and over for 18 years - the largest study yet on road traffic noise and Parkinson’s disease. [36] For every 11.5-decibel rise in noise at the loudest wall of a home, the risk of Parkinson’s rose about 3%. [36] The authors adjusted for air pollution and income. Even so, an outside expert noted the design “doesn’t prove causation” and called noise more likely “a general stressor.” [36]

Both are careful, and both lean on the same quiet phrase: linked to. Today’s lesson is about how much work that phrase hides.

A weight-loss drug that missed on the heart

Eli Lilly reported that its triple-hormone drug retatrutide hit the main goals of two more phase 3 obesity trials. [13] A phase 3 trial is the big, final-stage test that compares a drug against a dummy in large groups. Patients lost up to 22.6% of their body weight. [13] But that peak fell short of Lilly’s earlier trials, and the drug failed to prove it cut cardiovascular events like heart attacks and strokes. [13] Weight loss is the easy headline. Whether these drugs also protect the heart is the harder, more valuable question, and this readout left it open. [13]

Outbreaks, records and access

Public health supplied the sobering news. The United States has now recorded 2,318 confirmed measles cases in 2026, the most in over three decades, as vaccination rates slip. [75] In November a panel will decide whether the country keeps its status of having eliminated the disease - which Canada lost last year. [75]

In the Democratic Republic of the Congo, a fast-spreading outbreak of the rare Bundibugyo strain of Ebola has prompted remarkably quick drug and vaccine trials. [1] They were launched in the middle of the emergency. [1] And Merck agreed to license an experimental HIV prevention pill to generic makers across 129 countries, a move that could widen access in the places that need it most. [55]

02 · Lesson · why it matters

Why "linked to" almost never means "caused by"

A correlation can be perfectly real and still tell a false story, when a hidden third thing quietly moves both the cause you notice and the effect you fear.

Two headlines, one small phrase

This week two careful studies made the news. Coffee drinkers have healthier livers. People on noisier roads get Parkinson’s disease a little more often. Both are large, both are serious work. And both rest on the same two words: linked to.

That phrase sounds like a fact. It is really a flag. It signals that we have found a pattern, not yet a cause.

The lurking third thing

Here is the trap. Two things can rise and fall together not because one causes the other, but because a third thing drives them both. Statisticians call it confounding.

Picture the noisy road. A home on a busy road is not just louder. It sits in more exhaust, more traffic fumes, often cheaper housing, worse sleep. Noise is only the part you can hear. Any of its companions could be the real culprit, with noise just riding along.

So “noise causes Parkinson’s” might really be “the pollution that always travels with noise causes Parkinson’s.” The pattern is honest. The arrow from cause to effect is a guess.

The same shape, in your coffee

Coffee is no different. People who drink a lot of coffee are not a random slice of humanity. They may be wealthier, more active, less likely to drink heavily, more likely to see a doctor.

Any one of those could protect a liver on its own. If the healthy-liver crowd also happens to love coffee, the coffee gets the credit that belongs to the lifestyle. The link is real. Who is doing the causing stays an open question.

Why the honest researchers say so first

The best part of both studies is what the scientists admit. The noise team adjusted for air pollution and income, trying to strip the third things out. An independent expert still said plainly that the design “doesn’t prove causation.”

That is not a failure. It is honesty about a wall every study of this kind hits. You can correct for the lurking variables you can name and measure. You cannot correct for the ones you never thought to check. There is always another third thing you did not see.

How you actually break the tie

There is one clean way out, and medicine leans on it hard. Take a group of people and split them by a coin flip. Some get the thing, some do not.

Because the flip is random, all the lurking traits, the wealth, the exercise, the ones nobody named, scatter evenly across both groups. Whatever difference is left at the end cannot be the third thing, because the third things were shared out equally.

That is the whole reason a randomized trial outranks a hundred “linked to” studies. It is also why the FDA scientists this week wanted trials for those peptides, and why “it worked for me” sits far below “a trial showed it worked.”

The web we are all standing in

None of this means ignore the science. It means read it at its true weight. Almost nothing in a body happens for a single reason. The thing you can measure is tangled with a dozen you cannot, and the neat arrow in the headline was drawn by a human who knew that.

Every week a fresh “X linked to Y” arrives, and it is easy to quietly rebuild a life around the arrow. Someone who has seen the lurking variable holds that arrow more loosely. Not because the finding is worthless, but because a single correlation can only carry so much on its own.

03 · Lab · your turn

Find the lurking variable

Rehearse spotting a confounder by holding candidate variables steady and watching a correlation survive or collapse.

04 · Hope · carry this

It is slow and unglamorous to insist that a link be tested before it is believed. But that same patience is the tool that has turned a hundred hopeful correlations into real medicine, and it is still working.