An Irish analysis has just examined whether drinking fluoridated water throughout life is associated with worse cognition in old age. The most comfortable headline would be “fluoride does not damage memory.” But the study says something more precise and more interesting: in this population and with this way of measuring, this association was not detected. Distinguishing both phrases is the real learning.
The discovery that does not say what it seems to say
The work was published in the Journal of Dental Research. He analyzed 2,495 Irish adultswith a mean age of 65.3 years, and found no statistically significant association between greater cumulative exposure to water fluoride and worse performance on two common cognitive tests.
This result is reassuring, but it has a logical limit that should be set from the beginning: Not detecting an association is not the same as showing that there is no effect. A study may not find a relationship for many reasons, and only one of them is that a relationship does not exist. Reading this work well consists of not confusing “it was not found” with “it was proven that there is none.”
How half a century of tap water was rebuilt
The most ingenious part of the study is also the one that marks its limits. No one measured the fluoride inside the participants’ bodies. Instead, the researchers reconstructed likely exposure from where each person had lived. To enter the analysis, a residential history of at least 50 years was required.
The procedure was as follows: each participant’s previous addresses were geographically located, linked to maps of water supply areas, and each supply was then cross-referenced with historical monthly fluoride concentration records. Thus, an estimated annual concentration per household was obtained. Adding these concentrations to the time spent in each home, a measure of lifetime cumulative exposure was constructed.
It is a reasonable strategy to cover decades that no laboratory could recover today. But it works like a probable exposure map, not as an individual consumption record. Cognition, for its part, was evaluated with the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA), two screening tests that explore various mental functions, but do not replace a complete neuropsychological evaluation.
What the study can’t tell us
Estimating exposure from home implies leaving out a lot of actual intake. The method does not capture bottled or filtered water, tea and foods prepared with fluoridated water, toothpastes, supplements, or individual differences in the amount of water each person drinks. Two neighbors with the same tap water may have received very different doses.
Furthermore, it is an analysis observational and transversal: observes people as they are, without assigning exposures, and measures cognition at a specific moment in time. That design is vulnerable to so-called residual confusion. Factors such as education, socioeconomic status, vascular health, nutrition, healthcare access, smoking, or alcohol consumption can influence both estimated exposure and cognition, clouding any relationship.
There is one more statistical nuance. A null result may also appear when exposure varies little between participants, when there are measurement errors, or when the sample is not powered enough to detect small effects. Therefore, the honest formulation is “no association was detected”, not “any effect was ruled out”.

Habitual fluoridation and high exposures: two different questions
Here appears the most common confusion in fluoride debates. The Irish study examined cumulative exposure from routine community fluoridation in older adults. That’s a question. What happens with high concentrations or during brain development is quite another.
He National Toxicology Program (NTP) of the United States reviewed the evidence on fluoride and neurodevelopment and concluded that the body of data supports considering fluoride a neurocognitive hazard to humans, especially at high exposures. But that same review noted that the information is insufficient to clearly establish the effects of the level of fluoride commonly used in water fluoridation.
They are, therefore, two investigations that answer different questions and should not be added as if they were a single test on all doses and all ages. The Irish study looks back at older adults and everyday exposures; The NTP review looks primarily at development and high concentrations.
Who is missing in this photograph?
Every sample leaves people out, and knowing who is missing is key to interpreting the results. In this case, only near the 5% of participants had documented exposure to fluoridated water between ages 0 and 5 years. With such a low proportion, the study can hardly say anything about the possible effects on early brain development, which is precisely the stage that is of most concern in other research.
There is another important detail. The corresponding wave of the TILDA study started with a larger initial sample, but the final analysis stopped at 2,495 people because it required a residential history of at least 50 years, an identifiable water source for each period, and an available cognitive assessment. This filter could favor people with more stable life trajectorieswhich introduces a possible selection bias.
The implication is clear: the results describe older Irish adults capable of providing a complete history. They cannot simply be extrapolated to children, pregnant people or populations exposed to different environmental concentrations.

How to read this type of studies and what is left open
Beyond fluoride, this work is a good exercise in critical reading. What is established is concrete and limited: in this sample and with this method, no association was found between the estimated cumulative exposure and worse performance on the MMSE or the MoCA. That is a fact of the study, not a hypothesis.
What remains open is just as important. We do not know if there are subtle effects that these screening tests do not capture, or what happens in other age groups or with other doses. Studies with direct biological measurement of fluoride and greater representation of childhood exposure would be needed to answer these pending public health questions.
References: Sharma V, O’Connell B, Zgaga L, et al. Lifetime Fluoride Exposure and Cognition in Older Irish Adults. Journal of Dental Research. 2026;0(0). doi:10.1177/00220345261490925