Mind & Body · Sunday, 11 October 2026

01 Briefing what happened

The US food regulator's draft rule would let some chemicals be added straight into food without a safety review, if each stays under 0.5 parts per billion

A rule sold in August as closing one food-safety loophole also widens another, campaigners found. Elsewhere, US reporters counted over 1,000 eye-damage reports on weight-loss drugs, and a world survey measured trust in science.

0.5

parts per billion: the amount under which a chemical would skip review

That is half a gram in 1,000 tonnes of food, and the chemical must not cause cancer. [1]

99%

of chemicals in US food came in without a full FDA review

They came through GRAS, where companies vouch for their own ingredients. [2]

111

food chemicals or substances found using the GRAS route

A 2026 review of US records by the Environmental Working Group, a campaign group, and the scientist Maricel Maffini. [1]

people who lost their gallbladder after eating one untested ingredient

The FDA learned tara flour was in food only after hundreds of people fell ill. [2]

The lead story — what happened

The US Food and Drug Administration (FDA), which decides what may go into food and medicine, has proposed letting some chemicals into food with no safety review. [1]
Today an exemption called the threshold of regulation covers only chemicals in packaging and machinery that touch food. [1] To qualify, a chemical must not cause cancer and must stay under 0.5 parts per billion. [1]
The draft would stretch that exemption to chemicals added straight into food: flavourings, preservatives, emulsifiers, binders and other ingredients. [1]
The change sits inside the FDA's draft rule on GRAS, short for generally recognized as safe. [1] Robert F. Kennedy Jr., the US health secretary, presented that rule in August as closing a loophole. [1]
Under GRAS, a company's own scientists can declare a new ingredient safe and then simply tell the FDA, or not tell it at all. [2] NPR reported that 99% of chemicals in US food came in this way, not through a full review. [2]
Campaigners say some chemicals that do not cause cancer still harm the body below 0.5 parts per billion. [1] They name perchlorate, some phthalates, some bisphenols and some PFAS, the 'forever chemicals'. [1]
Bisphenols and phthalates can latch onto the body's hormone receptors, the switches that hormones normally turn on. [5]
Perchlorate shows what can happen. In 2005 the chemical company BASF used the exemption for big grain sacks, and FDA tests later found more perchlorate in children's cereal. [1]
Perchlorate stops the thyroid, a gland in the neck, from taking in iodine. [1] The thyroid needs iodine to make T4, a hormone the growing brain depends on. [1]
A review of food-safety science says some chemicals act at very low doses. [4] That challenges the usual idea that less of a chemical always means less harm. [4]
The draft also takes no account of several small doses adding up, campaigners say, and shoppers would have no way to know a product holds an exempted chemical. [1]
The FDA did not answer The Guardian's questions. [1] It is not known who added the exemption to the rule. [1]

In packaging, todayNo review

Added to food, todayReview

Added to food, draftNo review

What happens to a chemical kept under 0.5 parts per billion that does not cause cancer. The draft rule moves food ingredients into the no-review group.

Who is involved

The US Food and Drug Administration (FDA)

the US agency that decides what may go into food and medicine; it wrote the draft rule
Robert F. Kennedy Jr.

the US health secretary; he said in August the GRAS loophole was being closed
Maricel Maffini

an independent food consultant who has worked on the exemption; she says small amounts can still harm
BASF

a German chemical company; it used the exemption in 2005 for perchlorate in grain sacks

How it unfolded

2005 BASF uses the exemption for perchlorate in big grain sacks [1]
2017 FDA data show more perchlorate in children's cereal; six months later it approves the sacks again [1]
Jul 2026 The FDA turns down a petition for legal limits on PFAS in food [3]
Aug 2026 Kennedy and the FDA say the GRAS loophole is being closed [1]
10 Oct 2026 The Guardian reports the wider exemption inside the same draft rule [1]

Where this points

The next test is the final GRAS rule: whether the wider exemption stays in it, and whether the FDA says who added it. [1]

What is pushing on the whole day

The bar and the word are our reading of how hard each one is pushing today. The arrow is where it is heading. The evidence is in the stories below.

Harms found after sale Building↑

Perchlorate turned up in children's cereal in FDA tests years after it was let into grain sacks. [1] NPR counted over 1,000 reports of eye-nerve damage on weight-loss drugs since 2024, and the FDA has only now begun its own study. [6] Scientists are testing in fish embryos whether dolutegravir, an HIV drug, harms the developing heart. [12]

Patients and shoppers left uninformed High→

Shoppers would have no way to know a food holds an exempted chemical. [1] When Todd Engel lost the sight in one eye on Ozempic, his doctor said he had heard nothing about such a risk. [6] A mother whose son is held at Rampton, a secure hospital in England, says staff tell her almost nothing. [8]

Care moved onto phones and into villages Building↑

Weekly text messages went to 409 young people in Brazil to support HIV prevention and care. [11] A Hong Kong phone app finds plaque on children's teeth from six photos. [15] Quick blood tests for river blindness were tried in seven communities in Ghana. [10]

Daily habits showing in the body Steady→

In 46,843 British adults, an unhealthy lifestyle went with brains that looked older on scans. [9] In 325 young Polish men, heavy problem internet use went with worse health habits, such as eating. [13]

The rest of the day

11 more stories on this beat.

Each with its own sources. None of these is a link to the story above.

02
Over 1,000 eye-damage reports on weight-loss drugs

NPR, the US public radio network, counted more than 1,000 reports of optic nerve damage linked to GLP-1 drugs in the FDA's side-effect database since 2024. [6] That is out of nearly 190,000 reports about the drugs, mostly stomach upsets or dosing mistakes. [6] More than one American in ten takes a GLP-1 drug, such as Ozempic, to lose weight. [6] Doctors, patients and drug companies send reports by choice, and the FDA does not check them. [6] The FDA says it has started its own study of the eye risk. [6]

1,000+Optic nerve reports
190,000All reports
Reports to the FDA about GLP-1 drugs since 2024. Most are minor problems.

Why it matters — Two large US lawsuits now include 247 people with this eye damage and more than 4,000 with pancreatitis or gallstones. [6] Novo Nordisk, which makes Ozempic, says such harms are very rare in its trials and the cases have no merit. [6]
03
Americans trust science more than most

The Wellcome Global Monitor surveyed more than 140,000 adults in about 140 countries, with the polling company Gallup. [7] Wellcome is a British charity that pays for research. [7] It found 52% of Americans trust science 'a lot', against 26% across the world. [7] Worldwide, the share who say science helps most people in their country fell from 42% in 2018 to 37% in 2025. [7] In the US it rose from 48% to 57%. [7]

US52%
World26%
Share of adults who trust science 'a lot', Wellcome Global Monitor.

Why it matters — STAT, a US health news site, found the US far more divided than other rich countries: a 24-point gap between religious and non-religious people, against 5 points elsewhere. [7] Only 40% of Americans said they have a lot of confidence in their hospitals and clinics. [7]
04
A mother's view inside a secure hospital

England has three high-security hospitals for people with severe mental illness who are a grave danger to themselves or others: Rampton, Ashworth and Broadmoor. [8] The Guardian spoke to Jan Neate, whose son Jake has been held at Rampton for eight years. [8] In 2017, during psychosis, he killed his partner, Suzanne Brown. [8] His psychosis returned after clozapine, a schizophrenia drug that had kept him stable for ten years, was stopped suddenly. [8] A later report found no new risk check was done. [8]

Why it matters — Neate says she is told almost nothing during his spells of seclusion, when he is kept alone in a room. [8] England's health watchdog raised Rampton from 'inadequate' to 'requires improvement' in May 2025, but said staff shortages remained. [8]
05
Genes and habits both age the brain

Researchers at the Kunming Institute of Zoology in China used brain scans, genes and lifestyle data from 46,843 people in UK Biobank, a large British health study. [9] They built a model that guesses a person's age from their brain scan. [9] Higher genetic risk and an unhealthy lifestyle both went with brains that looked older than their owners. [9] The team found 19 places in the DNA linked to how fast the brain ages. [9]

Why it matters — At one of those places, people with the risky version aged faster on a poor lifestyle and slower on a healthy one. [9] The finding comes from watching people, not from a trial that changed their habits. [9]
06
New river blindness tests fall short

Scientists in Ghana tried two new quick tests for onchocerciasis, or river blindness, a disease caused by a parasitic worm. [10] They tested 1,700 people aged five and over in seven communities. [10] Some had been given mass medicine against the worm for years, and some were thought free of it. [10] A lab test on skin samples found the worm in only 26 people. [10] The new tests caught about seven cases in ten, and wrongly flagged about three in ten people without it. [10]

Why it matters — The tests were good enough to map where the disease is, by World Health Organization standards, but not to decide when mass treatment can stop. [10] With so few real cases, the authors say it is hard to judge any new test. [10]
07
Text messages kept young people on HIV care

Doctors at Fiocruz, Brazil's public health institute, offered 409 young gay, bisexual and non-binary people aged 18 to 24 a clinic service built for them, plus weekly text messages. [11] All 409 accepted, and 83% were still in the study after 48 weeks. [11] During the messages, more took HIV medicine and PrEP, pills that prevent infection, and binge drinking and depression scores fell slightly. [11] 99% said they would recommend the messages. [11]

Why it matters — There was no comparison group, so the study cannot show the messages caused the changes. [11] Young men who have sex with men carry a large share of HIV in Latin America, and services built for them are scarce. [11]
08
An HIV drug harmed fish embryo hearts

Scientists at the University of Brescia in Italy exposed zebrafish embryos to dolutegravir, a drug that treats HIV. [12] Earlier work linked the drug to problems in unborn babies, especially of the spine and brain. [12] At one dose, the fish embryos had weaker heartbeats, less blood flow and shrunken heart chambers. [12] The drug also changed genes that build the heart and blood vessels. [12]

Why it matters — The authors call the work exploratory, and fish embryos are not human babies. [12] Drugs of this kind have greatly cut deaths from HIV, so their effects on unborn babies are studied closely, as here. [12]
09
Heavy internet use and young men's habits

Researchers at Pomeranian Medical University in Poland studied 325 men aged 18 to 30. [13] About 18% had a high or very high level of problem internet use. [13] One in three led an unhealthy lifestyle. [13] Heavier problem use went with worse health habits, including eating, and weakly with lower serotonin, a brain and gut chemical, in the blood. [13]

Why it matters — The links were small, and the study measured everyone once, so it cannot say which came first. [13] The authors want help for young men to cover both screen use and daily habits. [13]
10
How the brain judges a group's size

Researchers scanned the brains of 29 adults while they judged the average size of groups of objects on a screen. [14] A nearby group of large shapes made the target group look smaller. [14] A group of small shapes made it look bigger. [14] Brain scans showed the change in the parts of the brain that first handle what the eyes see. [14] It is a group version of an old optical illusion. [14]

Why it matters — It shows the brain does not judge a group on its own; it compares it with the groups nearby. [14] Areas higher up in the brain tracked how strongly each person was fooled. [14]
11
A phone app spots plaque on children's teeth

Dentists at the University of Hong Kong built PlaqueSAM, a tool that finds plaque, the sticky film that causes tooth decay, from six phone photos. [15] It learned from 3,402 photos of 567 children in 11 kindergartens. [15] Tested on 81 children, it was right about 80% of the time. [15] It needs no dye, which dentists usually use to show plaque. [15]

Why it matters — Parents could one day check their child's brushing at home. [15] The authors say it still has to be tested in homes, with parents taking the photos. [15]
12
Low 'good' cholesterol and infections in liver disease

Doctors studied 311 people with cirrhosis, a scarred liver that no longer works well. [16] Those with low HDL, the 'good' cholesterol, or low levels of its main protein were more likely to need antibiotics for an infection within a year. [16] Infections hit about a quarter to a third of patients. [16] Among those who got an infection, 41% died during the study. [16]

Why it matters — A simple blood test might one day show who needs closer watching. [16] The study was in one hospital, and the authors say it must be checked elsewhere. [16]

02 Lesson why it matters

A rare harm shows up only after many thousands of people are exposed

A trial or a safety check sees a few thousand people, so a harm that hits one in 10,000 is usually found later, among the millions who use the product.

The twist

A safety test can only find a harm that happens inside it. A harm that hits one person in 10,000 is usually found among the public, after the product is on sale.

The picture

Trial of 3,000 people0.3

1 million users100

Made-up numbers. A harm that hits 1 in 10,000 people is expected 0.3 times in a trial of 3,000, so usually never, and 100 times among a million users.

How it works

A trial or a safety check studies a few thousand people
A harm that hits one person in 10,000 may not happen once in that group
The product goes on sale and millions of people use it
Now the rare harm happens dozens or hundreds of times
Scattered reports appear, and a new study must test whether the product caused them

The same force, elsewhere today

Where this chain is also running, in today's other stories.

Over 1,000 eye-damage reports on weight-loss drugs

More than one American in ten now takes these drugs, and only at that scale did over 1,000 eye-nerve reports pile up and the FDA start its own study.
The draft food-chemical rule

A chemical under 0.5 parts per billion would skip review, so any harm would first show up in people eating it, as perchlorate showed up in children's cereal years later.
New river blindness tests fall short

With only 26 real cases among 1,700 people, there were too few to tell how well the new tests work.

Where you've seen this

Car recalls

a fault in one car in 100,000 shows up only after a million are on the road

New phone software

a rare crash appears when millions of people install the update, not in the lab

Aircraft parts

a crack that forms once in many thousands of flights is found in service, not in testing

The catch

A pile of reports after sale is not proof. Many people taking a drug get ill for other reasons, so a careful comparison is needed before anyone knows the product caused it.

And the whole of it

Every person who eats a new ingredient or takes a new drug is part of the group where its rare harms are finally found. Each of them sees only their own body, and the regulator sees only the reports someone chose to send.

03 Truth what's really going on

What is really going on

The FDA presented its draft GRAS rule in August as closing a food-safety loophole. The same draft also lets chemicals kept under 0.5 parts per billion be added to food with no review, and the agency did not say so or answer questions about it. [1] On the rest of the day, harms are being counted after products reach people: NPR found over 1,000 eye-nerve reports on weight-loss drugs before the FDA began its own study. [6]

Why it works on us — Half a gram in 1,000 tonnes of food sounds too small to matter, so few people ask what a hormone-like chemical does at that amount.

Who gains

Food and chemical companies — Under the draft they could add chemicals kept under 0.5 parts per billion as flavourings or preservatives without waiting for an FDA review. [1]
Companies that vouch for their own ingredients — GRAS lets their own scientists declare a new ingredient safe and simply tell the FDA, or not tell it. [2]
Makers of GLP-1 drugs — The FDA has not asked for new eye-risk warnings on US labels while its study runs, unlike in Europe, lawyers say. [6]
Young people in Brazil's Conectad@s study — A clinic service built for them, with weekly texts, kept 83% in care for 48 weeks. [11]

Who pays

Children who eat cereal — FDA tests found more perchlorate, which blocks the thyroid's use of iodine, in children's cereal after the 2005 exemption. [1]
Shoppers in the US — They would have no way to know whether a food holds a chemical that skipped review. [1]
People who lost sight on weight-loss drugs — Todd Engel kept taking Ozempic until his second eye failed, because neither he nor his doctor knew of the possible link. [6]
Families of patients in secure hospitals — Jan Neate says she is told almost nothing during her son's seclusion at Rampton. [8]
Ghana's river blindness programme — The new tests are not yet good enough to decide when mass treatment can safely stop. [10]

What nobody knows yet

Open questions from across today’s stories — ours included.

01

Who added the wider exemption to the draft rule.

The Guardian says the US president's office or the US budget office took part in the talks on the rule, and it is unclear who added the change. [1]
02

Which chemicals would qualify, and how many products would carry them.

The draft sets a test of 'no appreciable risk' and names no list, and shoppers would have no way to see an exempted chemical on a label. [1]
03

Whether GLP-1 drugs cause the eye-nerve damage reported by patients.

Reports to the FDA are voluntary and unchecked, the main study found a link but not a cause, and the FDA's own study has no published result. [6]
04

Why Americans trust science more than most countries while being more divided over it.

The Wellcome survey does not explain the gap, and one expert told STAT that Americans tend to answer surveys at the extremes. [7]
05

Whether Jake Neate will ever move to a lower-security hospital.

He is held on an indefinite order that needs the UK Home Office to agree, and the trust says the decision takes time. [8]
06

Whether the new river blindness tests are accurate enough to use.

Only 26 people in the study really had the worm, so the estimates of how well the tests work are wide. [10]
07

Whether dolutegravir harms the hearts of human babies.

The heart damage was seen in zebrafish embryos at one dose, and the authors call the work exploratory. [12]
08

Whether changing habits would slow brain ageing in people with the risky gene version.

The UK Biobank study watched people as they were and did not test a change in lifestyle. [9]

04 Hope carry this

In the United States, the share of adults who say science helps most people in their country rose from 48% in 2018 to 57% in 2025.

Also true today

In a Brazilian study, 409 young gay, bisexual and non-binary people were offered a clinic service built for them and weekly text messages. All of them said yes, and 83% were still taking part after 48 weeks.
Among 46,843 British adults, people carrying a gene version that sped up brain ageing on an unhealthy lifestyle had slower brain ageing on a healthy one.