The Fluoride Limit Was Written to Stop Crippling Bone Disease

What was that limit written to stop?

Bone disease. Not teeth. Not brains. Bone.

Picture the far end of it. Bone goes weak and brittle and breaks more easily. Soft tissue near the joints turns hard. A shoulder that used to swing stops partway, and then it stays there. That’s stage III skeletal fluorosis, and a footnote in this summer’s plan uses a blunter name for it, the crippling stage.

Somebody was protecting you from that in 1986. They set a ceiling on fluoride in public tap water at 4 parts per million, and it hasn’t moved since. Go over 4 and whoever runs your pipes has to fix it. Stay under, and nobody has to do a thing.

Forty years is a long run for a limit on your drinking water. Its age isn’t the interesting part, though. What it was aimed at is.

Because a ceiling is written against one harm, and everything it was never written against falls outside it by design rather than by oversight. That is true of every number in this piece, and it is the reason a limit can be both correct and beside the point at the same time.

Was the limit set with children’s brains in mind?

They couldn’t have. That question had barely been put.

Stand at your kitchen tap for a second with a kid in the room. What you want to know is whether what comes out of it does anything to how that kid’s brain develops. Crippling bone disease and brain development are two different questions, and a figure built to stop one was never designed to answer the other.

Here’s the sentence doing most of that work in 2026, and you’re getting it secondhand on purpose. Summarizing a 2024 federal review, EPA writes that fluoride exposure was associated with lower IQ in children, with uncertainty in that relationship below 1.5 parts per million. That summary is what we opened. The review underneath it we did not, so take that finding as relayed.

We think the gap between those two questions carries the whole story, and most coverage of this lost it in a first paragraph. A ceiling written for one endpoint is being checked against another. Nobody has said it failed.

Has EPA decided anything yet?

EPA published a plan on 3 August 2026. Not a result. A plan.

It sets out how that check will be run, and it came out before any of the work, which is the entire point of writing one. Publish your method first and you can’t quietly build a study around whatever answer you wanted.

So nothing you drink changed on 3 August. Same ceiling in August as in July.

Somebody had to go through the science first, and the size of that job is worth a beat. Over 260,000 studies got pulled. AI tools ranked over 70,000 of those, and screening left over 500 that bear on two effects in children, stained and pitted teeth and brain development.

One thing about scope, though, and it cuts both ways. This review looks at harm from fluoride and not at benefit. Fluoride at lower exposure cuts tooth decay, which that same plan calls “one of the most common chronic diseases among American children,” and none of that gets weighed here. That’s the sort of thing we’d want flagged before a result lands rather than after. Read whatever comes out of it knowing half a ledger was left off on purpose.

Four steps run this process, and August was step two. A draft comes next, out for public comment and outside review, then a final document. A normal clock would have left none of it due until 2030. Lee Zeldin, who runs the agency, said this milestone means “we are another step closer to completing this assessment years ahead of schedule.”

Has anybody done this arithmetic before?

And then it went in a drawer.

An earlier round of review had the agency work out what a revised health goal for fluoride might be. It got 0.9 parts per million. Then it concluded that the rule wasn’t a candidate for revision at that time, pointing at higher priority work and at emerging research on brain development.

Set 0.9 beside the 4 still standing over your tap.

We’d read that twice, because both halves can be reasonable at once. Working out a figure on paper is a technical job. Reopening a national drinking water rule is a much bigger one, with cost and feasibility and every utility that would have to change.

Scientists outside the agency had pushed in a similar direction even earlier, and the same paperwork records it. In 2006 a National Research Council panel went through the health data and recommended lowering that 4. Their grounds were that it doesn’t protect against severe staining and pitting of teeth, and may not protect against broken bones.

So this review doesn’t start from nothing. It starts with all of that already on the record.

What does your water company have to tell you?

If your water goes over 2 parts per million, a letter has to arrive.

Not a fine. Not a shutdown. A letter. Federal rule fixes that trigger at 2, well under the enforceable 4, and then does something rules rarely bother with. It writes your utility’s first sentence for them.

“This is an alert about your drinking water and a cosmetic dental problem that might affect children under nine years of age.”

That notice goes out within 12 months of a system learning it’s over, and again every year while your water stays that way. Copies go to every new customer as service starts, and to your state public health officer. It tells households to give children under nine a different source of drinking water, or the same supply with fluoride taken out of it. Older children and adults, it says, may safely drink what’s there.

Notice what it names. A cosmetic problem, staining and pitting on teeth, rather than anything deeper. That’s the job 2 was set to do, and it’s a different job from 4.

If your water has ever run over 2, one of those letters should have reached you. We’d go and look for it.

Who decides whether your town fluoridates?

Least of all the agency running this review. Law bars it from both.

Say your town votes fluoride out next spring, or votes it back in. Nobody federal gets a say in that. EPA sets a ceiling on what may be in your water, and under the Safe Drinking Water Act it can’t require anybody to add a substance for preventive health care. It makes no recommendation about fluoridating, either way. That call belongs to your state and to whoever runs your pipes.

A separate federal figure covers towns that do add it. The Public Health Service recommended 0.7 parts per million in 2015, and that’s advice rather than a limit.

Three figures, three jobs, and mixing them up is how two people end up arguing past each other over one fact.

That ceiling also doesn’t care how fluoride got into your glass. Plenty of water carries it naturally, straight out of whatever rock it ran through, and about 11 million people were served by systems above 0.7 with nobody adding a thing. That figure dates from 2015.

What if you are on a private well?

No ceiling. No letter. No annual notice. Nobody checking.

EPA counts around 15 percent of this country drinking from a private well, over 43 million people. If that’s you, your well stands outside the Safe Drinking Water Act, and outside most state law as well. Wells serving fewer than 25 people are written out of that act by name.

Fluoride in groundwater comes out of rock, so how much yours carries depends on what it traveled through on its way up to you. Nothing in this review changes that, and nobody’s going to knock on your door about it.

The practical consequence cuts both ways, which is why it is worth stating. A private well can sit far under any of these figures, and plenty do. It can also run far over them, in parts of the country where the rock is generous with it, and the owner has no letter, no annual notice and nothing to tell them so.

So a fluoride result is one more line on the panel you were already going to run. Ask for it by name when you send a sample, because it isn’t in the standard four, and a lab won’t add it unless you say so.

What should you do while this plays out?

Read one line on a page that already comes to your house.

Every community water system has to mail out an annual report on what’s in your tap, due by July 1, and fluoride’s on it. Find your figure, then line it up against those three. If yours never turned up, call whoever bills you and ask.

Your tap isn’t all of it, either. Somewhere between 40 and 70 percent of what an average person here takes on arrives through drinking water, and the rest comes through food, toothpaste, mouthwash and supplements. That is worth knowing before you rebuild a household around one figure on a page.

Keep the timing straight too. Public comment on an earlier plan closed in February 2026, and August’s plan opened no new window. A draft goes out for comment when it lands, and that’s the moment worth showing up for.

One last thing, and we’d keep it well away from everything above. Nothing has been concluded. No limit has moved. Somebody wrote down how they plan to do a piece of work, and reading that as a verdict is how most of the noise around this got started. What happened in August is that a ceiling set in 1986 against crippling bone disease was formally pointed at a question nobody was asking in 1986. No date on the answer.