Fluoride, to me, was always a teeth story. You spit, you rinse, you move on. So when a paper landed this week tying the fluoride in kids’ drinking water to their blood pressure and cholesterol, I squinted at it, and then squinted harder when I got to the dose. The reassuring headline it will earn rests on tap water carrying less fluoride than US public-health guidance tells your town to add.

The study is small and clean. Researchers in Yazd, Iran recruited 161 children and teenagers, ages 6 to 18, and did the thing most fluoride studies skip: they measured the actual dose. They sampled the water at 56 points across the city’s supply, then measured what was coming out of each child in their urine, which is where fluoride ends up once the body is done with it. Then they lined that up against a full cardiometabolic panel, blood pressure, LDL, HDL, triglycerides, total cholesterol, fasting glucose, magnesium, calcium. The workup you’d order if you were worried a child’s heart was already in trouble.

The headline finding is that there was no finding. Once the researchers corrected for urine dilution using creatinine, no significant association turned up between a child’s fluoride level and any of those heart-risk markers. If you want a reason to relax about the fluoride in your kid’s water, this is the paper someone will hand you.

So read the dose before you exhale. The children in Yazd carried a mean urinary fluoride of 0.49 mg/L, and their tap water averaged 0.53 mg/L. That sits below the 0.7 mg/L the US Public Health Service still recommends communities add to their water. This study didn’t test the high-exposure, above-target water people are actually fighting about. It tested a low dose and found, reassuringly, that a low dose behaves like a low dose.

DOSE VS US TARGET
US recommended level0.7 mg/L0.53 mg/LYazd tap water
The reassuring null was measured below the level US communities are told to add. Source: US Public Health Service, 2015

The question the reassuring headline skips is why fluoride would reach the heart at all. A mineral we file under enamel has no obvious business in a blood vessel. But follow the exposure up the dose curve and a mechanism starts to sketch itself. In a study of children in Chihuahua, Mexico, where many kids carried far more fluoride, urinary fluoride tracked with rising VCAM-1 and ICAM-1, the adhesion molecules a vessel wall switches on when it is inflamed, and with thicker carotid artery walls. The authors read those as vascular-injury markers, the early measurable signs that a vessel is under stress, not proof of disease. That is enough to float a plausible mechanism: at high enough exposure, fluoride may irritate the endothelium rather than simply pass through it.

A second Chihuahua study genuinely surprised me. When researchers put ultrasound on 374 children’s hearts, the ones in the highest fluoride tertile carried more left ventricular mass, about 3.97 grams more muscle in the heart’s main pumping chamber than the lowest-exposure kids, along with higher cardiac output. A child’s heart bulking up is not a sign of fitness. It is a heart working against something. Wait, why would a mineral do that to muscle it never directly touches? We don’t know yet, but a heart adding wall like that is responding to strain, and the strain is showing up right alongside the fluoride.

EXTRA HEART MUSCLE
3.97 gramshighest vs lowest fluoride tertile
More left ventricular mass in 374 Chihuahua children, by urinary fluoride. Source: Environmental Research, 2025

Zoom out to a whole population and the direction holds. In Changwu Town, China, where water fluoride has historically run a brutal 1.2 to 4.5 mg/L, one study found stroke prevalence climbing step by step with cumulative fluoride exposure, a dose-response that reached statistical significance. That is an ecological study, the weakest design in the pile, and I’ll say it plainly: it can’t prove fluoride caused a single stroke in that town. But it rhymes with the studies underneath it.

So look at what stacks up, and watch the dose. At the low Yazd exposure, kids’ hearts looked fine. At higher exposure, the Chihuahua children showed vascular-injury markers and heavier heart muscle, and one Chinese town logged more strokes as exposure rose. It is the same shape the federal government’s own scientists found for the brain: the National Toxicology Program concluded with moderate confidence that higher fluoride exposure is associated with lower IQ in children, while admitting there simply wasn’t enough data to judge the 0.7 mg/L level Americans actually drink. Harm at the high end, a shrug at the dose in your kitchen. That is the exact gap a low-fluoride town like Yazd cannot fill, no matter how carefully it measures.

Credit where it is due. The Yazd work wasn’t industry-funded. It came from the Iranian National Institute for Medical Research Development and a state university, with the authors declaring no competing interests. This isn’t a chemical-lobby product dressed up as reassurance. It is a careful null. Which is exactly why the wrinkle its own authors flagged matters so much. When they standardized fluoride to creatinine a different way, the numbers flipped and showed inverse associations with diastolic pressure, cholesterol and HDL. The authors say it in the paper: the whole result is sensitive to how you correct for how dilute a child’s urine is. A finding that swings on which dilution formula you pick isn’t an all-clear. It is a measurement warning sitting on top of a dose too low to move much in the first place.

This lands in the middle of a real fight, not an academic one. RFK Jr. has said he’ll direct the CDC to stop recommending fluoridation, the EPA has opened a fresh review of the risks, and Utah became the first state to ban it outright. The establishment reflex has been to wave studies like this one as proof the skeptics are overwrought. But Yazd’s own province undercuts that: drinking water in nearby Ardakan has run to 2.92 mg/L, more than five times what these particular children drank. The dose that harms is not hypothetical. It just wasn’t the dose this study looked at.

So here’s my conclusion. I’m not going to treat one careful null at a near-nonexistent dose as the last word on fluoride and the heart, because the higher-exposure studies keep pointing one way and the mechanism is starting to make biological sense. I drink tap water and I’m keeping my toothpaste. But I’d pull the actual fluoride number for my own water district before calling it safe, and if I were raising a kid in a county sitting well above 0.7 mg/L, I would not wait for the establishment to decide the dose in my child’s glass is finally worth measuring properly.

Sources

  1. Environmental Health – Relationship between urinary and water fluoride with cardiovascular risk factors in children and adolescents in Yazd (2026)
  2. Environmental Research – Vascular and kidney injury biomarkers in Mexican children exposed to inorganic fluoride (2019)
  3. Environmental Research – Left ventricular mass and systolic function in children environmentally exposed to fluoride (2025)
  4. Fluoride exposure from drinking water and stroke risk, Changwu Town, China (2024)
  5. National Toxicology Program – Fluoride exposure, neurodevelopment and cognition monograph (2024)
  6. US Public Health Service – Recommendation for 0.7 mg/L fluoride in drinking water (Federal Register, 2015)
  7. CNN – HHS and EPA to study fluoride as RFK Jr. moves to end CDC recommendation (2025)
  8. Deseret News – Utah becomes first state to ban fluoride in drinking water (2025)
  9. Fluoride concentration and health risk in drinking water, Ardakan, Yazd province, Iran