Sodium Fluoride

What it is

Sodium fluoride is a simple inorganic salt of sodium and the fluoride ion. It is the fluoride source in most toothpastes, in mouthrinses, in professionally applied and prescription dental gels, in swallowed tablets and drops, and it is one of the compounds used to fluoridate public water supplies. A prescription 1.1% (w/v) neutral sodium fluoride gel contains 5 mg of fluoride ion per gram, and the label notes that a thin ribbon of it contains about 2 mg of fluoride. The same label calls it a self-topical product for caries prevention, says it is not a dentifrice, and warns it is not for systemic treatment and should not be swallowed.

What the research says

For preventing tooth decay the evidence is strong and quantified at the low end of exposure, and it gets contested as exposure rises. A Cochrane review of 96 randomised trials rated it high to moderate certainty that toothpaste at 1000 ppm fluoride or above reduces decay, with a dose-response up to about 1500 ppm and nothing gained beyond it. For water fluoridation the 2024 Cochrane update found that in post-1975 studies the benefit shrank to about a quarter of one tooth, on low-certainty evidence whose confidence interval includes no benefit. On harms the picture splits by dose: at 0.7 ppm about 12% of people have dental fluorosis of cosmetic concern and 40% have fluorosis of some degree; above roughly 1.5 mg/L a dose-response meta-analysis finds rising fracture risk, and the US National Toxicology Program concluded with moderate confidence that higher exposures are consistently associated with lower IQ in children. That IQ conclusion is disputed, and the same team's meta-analysis found the drinking-water association null below 1.5 mg/L. As a drug for osteoporosis, at 75 mg a day, sodium fluoride increased bone density and tripled non-vertebral fractures.

Evidence grade: Well established.

How it works

Drug class: Inorganic fluoride salt; topical and systemic anti-caries agent

Fluoride works chiefly at the surface of the tooth, after the tooth has come through, rather than by being built into it from the bloodstream. Present at low concentration in saliva and plaque, it slows the dissolving of enamel by acid and speeds its repair, and it forms calcium-fluoride-like deposits that release fluoride back when the mouth turns acid. It also inhibits the enzymes and proton pumps of the bacteria that make that acid. High-concentration products applied frequently increase the tooth's resistance to acid dissolution and drive fluoride into the enamel. Because the action is local and reversible, it depends on repeated exposure rather than on a body store. (Source 1)

What it is used for

• This is the use with the strongest evidence. A Cochrane review of 96 randomised trials found high- to moderate-certainty evidence that toothpaste at 1000 ppm fluoride or more reduces decay, with a dose-response up to about 1500 ppm and no further gain above that. Prescription-strength products such as 1.1% sodium fluoride gel are for people at high risk. Evidence: established. (Source 2)
• A Cochrane review of 11 trials in 7,196 children found supplements were associated with 24% less decay in permanent teeth than no supplement, but the same review rated 10 of the 11 trials at unclear risk of bias and found no advantage over topical fluoride. The effect on baby teeth was unclear. Evidence: limited. (Source 3)
• This is a public-health water treatment rather than a licensed medicine use, so it sits outside any drug label. The 2024 Cochrane update found that in studies done after fluoride toothpaste became widespread, starting water fluoridation was associated with about a quarter of one tooth less decay, on low-certainty evidence whose confidence interval includes no benefit at all. The same review estimated 12% of people have fluorosis of cosmetic concern at 0.7 ppm. Evidence: disputed. (Source 4)
• Sodium fluoride at 75 mg a day was tried for osteoporosis in the 1980s. It raised spinal bone density by 35% and yet tripled non-vertebral fractures, because the new bone it builds is mechanically weaker. It was abandoned. This is the clearest demonstration that a bone density reading on fluoride does not mean stronger bone. Evidence: not-supported. (Source 5)

Interactions

• Calcium, milk and dairy products (label): Calcium binds fluoride in the stomach and gut so less is absorbed. The label puts this to work: for a small accidental swallow it tells you to give calcium, such as milk, and for a bigger one soluble calcium such as calcium gluconate or calcium lactate solution. The same chemistry means a calcium supplement or a glass of milk taken with a swallowed fluoride supplement will reduce how much fluoride gets in. (Source 6)
• Fluoridated drinking water and other fluoride sources (label): Fluoride adds up across sources - water, toothpaste, mouthrinse, supplements and tea. The prescription gel label flags the problem directly: swallowing it repeatedly can cause dental fluorosis in children under 6, especially where the water supply already exceeds 0.6 ppm fluoride. This is why dentists ask about water fluoridation before prescribing fluoride supplements. (Source 7)
• Food and drink straight after application (label): Eating, drinking or rinsing washes a topical fluoride product off the teeth before it has done its job. The label asks adults to spit out the gel and then not eat, drink or rinse for 30 minutes. (Source 8)
• Tea (theoretical): The tea plant concentrates fluoride in its leaves: a 2026 study of Camellia sinensis grown in solution describes it as a hyperaccumulator of both aluminium and fluoride, and the authors note that excess aluminium and fluoride pose health risks to humans. That is plant physiology, not a measurement in people - this study did not measure how much fluoride tea drinking delivers, and we found no human study of that in this search. So tea is a plausible dietary contributor to total fluoride intake, with no quantified human figure behind it here. (Source 9)
• A mouthpiece tray left in too long (label): Not a drug interaction but a product-handling one worth knowing: the label warns that prolonged exposure of the gel in a mouthpiece application, longer than one minute, may cause oral irritation such as burning. (Source 7)

Stopping it

• Nobody becomes dependent on fluoride and there is no withdrawal. But its benefit depends on continued exposure, because the mechanism is a local, post-eruptive one at the tooth surface: a 2026 mechanism review concludes the value rests on frequent, low-level exposures that keep the tooth in a repair-favouring state. Stop the exposure and the balance tips back. (Source 1)
• At the population level, what happens when fluoridation stops has barely been studied. The 2024 Cochrane update found a single eligible contemporary study and could not determine the effect of cessation on decay. (Source 10)
• For the swallowing-related harm, timing matters more than stopping: dental fluorosis is an enamel defect laid down while teeth are forming, which is why the prescription gel is contraindicated under age 6 unless a dentist or doctor recommends it. (Source 7)

What goes wrong

Harm from toothpaste itself was barely studied, and where it was reported it was minor. (Source 11)

• Systematic review, Low certainty.
• Size: a minority of the 96 included studies.
• Who: children, adolescents and adults.
• How long: mostly 36 months.
• Result: no rates given; reported effects were soft tissue damage and tooth staining, described as minimal.
• Funding: not stated in the abstract we read.

Only a minority of studies assessed adverse effects of toothpaste. When reported, effects such as soft tissue damage and tooth staining were minimal.

At the concentration used for water fluoridation, about 12% of people had dental fluorosis judged to be of cosmetic concern and about 40% had fluorosis of any degree. (Source 10)

• Systematic review, Low certainty.
• Size: 40 studies and 59,630 participants for fluorosis of aesthetic concern; 90 studies and 180,530 participants for any fluorosis.
• Who: populations exposed to differing water fluoride concentrations.
• How long: cross-sectional and other designs with concurrent control.
• Result: at 0.7 ppm, fluorosis of aesthetic concern approximately 12% (95% CI 8% to 17%); fluorosis of any level approximately 40% (95% CI 35% to 44%)
• Funding: not stated in the abstract we read.

With a fluoride level of 0.7 parts per million (ppm), approximately 12% of participants had fluorosis of aesthetic concern (95% CI 8% to 17%; 40 studies, 59,630 participants), and approximately 40% had fluorosis of any level (95% CI 35% to 44%; 90 studies, 180,530 participants).

The US National Toxicology Program concluded, with moderate confidence, that fluoride exposure above the WHO drinking-water guideline of 1.5 mg/L is consistently associated with lower IQ in children. (Source 12)

• Systematic review, Moderate certainty.
• Size: 72 studies of fluoride and IQ in children, of which 19 were rated high quality.
• Who: children; separately, adults.
• How long: cross-sectional and prospective cohort designs.
• Result: 18 of the 19 high-quality studies reported an inverse association between estimated fluoride exposure and children's IQ; 46 of the 53 low-quality studies also did.
• Funding: US National Toxicology Program (a US federal interagency programme)

Limit of this finding: Two things this conclusion does not say. It is about estimated exposures above the World Health Organization drinking-water guideline of 1.5 mg/L; the monograph does not draw a conclusion about the lower concentrations used in water fluoridation, and it ends by saying more studies are needed to understand whether lower exposure affects children's IQ. And most of the underlying literature is weak: of the 72 studies of fluoride and children's IQ, only 19 were rated high quality, and 46 of the 53 low-quality studies also reported an inverse association. These are associations in observational studies, not a demonstration that fluoride lowers IQ.

This review finds, with moderate confidence, that higher estimated fluoride exposures (e.g., as in approximations of exposure such as drinking water fluoride concentrations that exceed the World Health Organization Guidelines for Drinking-water Quality of 1.5 mg/L of fluoride) are consistently associated with lower IQ in children.

A dose-response meta-analysis found fracture risk rising with fluoride in drinking water above about 1.5 mg/L, and in women over 50 from as low as 0.5 mg/L. (Source 13)

• Meta-analysis, Low certainty.
• Size: 37 studies published 1945-2024 (19 on fractures, 10 on bone density, 8 on both)
• Who: general populations exposed to fluoride in drinking water; stratified by sex, age and fracture site.
• How long: observational studies of varying length.
• Result: fracture risk ratios of 1.06, 1.19 and 1.35 at 2.0, 3.0 and 4.0 mg/L compared with no exposure; in women over 50, fragility fracture risk ratio 1.26 at 1.0 mg/L; bone density associations inconsistent.
• Funding: not stated in the abstract we read; registered PROSPERO CRD42022321899.

We found a non-linear, positive relation between fluoride exposure and fracture risk, with an indication of an approximate threshold around 1.5 mg/L of fluoride in drinking water and a nearly linear increasing fracture risk above that concentration (risk ratios of 1.06, 1.19 and 1.35 at 2.0, 3.0 and 4.0 mg/L, respectively, compared to null exposure). Sex-specific dose-response analyses, available only for fragility fractures, indicated an effect in females but little or nothing in males. The association between fluoride and BMD was inconsistent, showing opposite trends for different bone sites (hip and spine) and amounts of exposure, and by sex. Among females aged over 50 years, an association of drinking water fluoride with fragility fracture risk started as early as around 0.5 mg/L (risk ratio of 1.26 at 1.0 mg/L).

When sodium fluoride was tried as a treatment for osteoporosis it increased bone density but increased fractures - the clearest demonstration that fluoride-laden bone is weaker bone. (Source 5)

• Randomized trial, Moderate certainty.
• Size: 202 postmenopausal women randomised; 66 fluoride and 69 placebo completed.
• Who: 202 postmenopausal women with osteoporosis who already had vertebral fractures, randomly assigned to sodium fluoride 75 mg per day or placebo, all given a calcium supplement of 1,500 mg per day.
• How long: 4 years.
• Result: lumbar spine bone mineral density rose 35% and femoral neck 12% (both P < 0.0001), but radius shaft density fell 4% (P < 0.02); new vertebral fractures 163 versus 136 (not significant); non-vertebral fractures 72 versus 24 (P < 0.01); side effects needing dose reduction in 54 versus 24 women.
• Funding: not stated in the abstract we read.

Limit of this finding: This was sodium fluoride used as a drug, not as a dietary or drinking-water exposure: 75 mg per day for four years, with a 1,500 mg per day calcium supplement in both arms, given to 202 postmenopausal women who already had osteoporosis and vertebral fractures. Only 66 of the fluoride group and 69 of the placebo group finished, so roughly a third of each arm dropped out. The figures are counts of events, not numbers of patients, and the paper gives no rates, so a per-woman risk cannot be worked out from them. Nothing here transfers to fluoride in toothpaste or in drinking water, where the exposures are orders of magnitude smaller.

The number of new vertebral fractures was similar in the treatment and placebo groups (163 and 136, respectively; P not significant), but the number of nonvertebral fractures was higher in the treatment group (72 vs. 24; P less than 0.01).

At the 75 mg a day dose used in that osteoporosis trial, more than a quarter of the women had side effects bad enough to need the dose cut - mostly stomach upset and leg pain. (Source 5)

• Randomized trial, Moderate certainty.
• Size: 202 postmenopausal women randomised; 66 fluoride and 69 placebo completed the trial.
• Who: 202 postmenopausal women with osteoporosis who already had vertebral fractures, randomly assigned to sodium fluoride 75 mg per day or placebo, all given a calcium supplement of 1,500 mg per day.
• How long: 4 years.
• Result: 54 of the fluoride group versus 24 of the placebo group needed dose reduction; main effects were gastrointestinal symptoms and lower-extremity pain.
• Funding: not stated in the abstract we read.

Limit of this finding: This was sodium fluoride used as a drug, not as a dietary or drinking-water exposure: 75 mg per day for four years, with a 1,500 mg per day calcium supplement in both arms, given to 202 postmenopausal women who already had osteoporosis and vertebral fractures. Only 66 of the fluoride group and 69 of the placebo group finished, so roughly a third of each arm dropped out. The figures are counts of events, not numbers of patients, and the paper gives no rates, so a per-woman risk cannot be worked out from them. Nothing here transfers to fluoride in toothpaste or in drinking water, where the exposures are orders of magnitude smaller.

Fifty-four women in the fluoride group and 24 in the placebo group had side effects sufficiently severe to warrant dose reduction; the major side effects were gastrointestinal symptoms and lower-extremity pain.

Swallowing a large amount of a fluoride dental product causes an acute poisoning picture within half an hour. (Source 6)

• Official position, Certainty not rated.
• Size: not given.
• Who: anyone, especially children.
• How long: symptoms may persist for 24 hours.
• Result: burning mouth, sore tongue, nausea, vomiting, diarrhoea, salivation, vomiting of blood, epigastric cramping and abdominal pain; the label's thresholds for action are 5 mg fluoride/kg and 15 mg fluoride/kg of body weight.
• Funding: manufacturer's own label (Westminster Pharmaceuticals)

Accidental ingestion of large amounts of fluoride may result in acute burning in the mouth and sore tongue. Nausea, vomiting, and diarrhea may occur soon after ingestion (within 30 minutes) and are accompanies by salivation, hematemesis, and epigastric cramping and abdominal pain. These symptoms may persist for 24 hours.

A high-fluoride prescription gel carries its own fluorosis warning for children under six, because they swallow it. (Source 7)

• Official position, Certainty not rated.
• Size: not given.
• Who: children under 6 years using a 1.1% sodium fluoride gel.
• How long: prolonged daily use.
• Result: no rate given; the label contraindicates use under 6 years unless a dentist or physician recommends it, and flags higher risk where water fluoridation exceeds 0.6 ppm.
• Funding: manufacturer's own label (Westminster Pharmaceuticals)

Prolonged daily ingestion may result in various degrees of dental fluorosis in pediatric patients under 6 years, especially if the water fluoridation exceeds 0.6 ppm, since younger pediatric patients frequently cannot perform the brushing process without significant swallowing.

What the evidence supports

Fluoride toothpaste at 1000 ppm or more reduces tooth decay compared with non-fluoride toothpaste, and the certainty of that is high to moderate. (Source 2)

• Systematic review, High certainty.
• Size: 96 randomised trials published 1955-2014; 81 trials in the network meta-analysis of permanent teeth; 55 studies for the 1000-1250 ppm comparison.
• Who: children, adolescents and adults.
• How long: follow-up in most studies 36 months.
• Result: standardised mean difference -0.28 (95% CI -0.32 to -0.25) for 1000-1250 ppm versus non-fluoride toothpaste, and -0.36 (95% CI -0.43 to -0.29) for 1450-1500 ppm; in young children's primary teeth, 1500 ppm reduced caries increment by 1.86 decayed/filled surfaces (95% CI -2.51 to -1.21)
• Funding: not stated in the abstract we read.

there was high- and moderate-certainty evidence that 1000 to 1250 ppm or 1450 to 1500 ppm fluoride toothpaste reduces caries increments when compared with non-fluoride toothpaste (SMD -0.28, 95% CI -0.32 to -0.25, 55 studies; and SMD -0.36, 95% CI -0.43 to -0.29, four studies)

Sodium fluoride supplements - tablets and drops - were associated with about a quarter less decay in permanent teeth, but on weak trials. (Source 3)

• Systematic review, Low certainty.
• Size: 11 studies, 7,196 children (3 studies for the comparison against no supplement)
• Who: children under 16 at the start.
• How long: minimum 2 years of follow-up.
• Result: 24% reduction in decayed, missing and filled surfaces (95% CI 16 to 33%); no differential effect against topical fluorides; effect on primary teeth unclear.
• Funding: not stated in the abstract we read.

In permanent teeth, when fluoride supplements were compared with no fluoride supplement (three studies), the use of fluoride supplements was associated with a 24% (95% confidence interval (CI) 16 to 33%) reduction in decayed, missing and filled surfaces (D(M)FS).

What the evidence does not support

Pushing the fluoride concentration far above 1500 ppm did not buy more protection, and the certainty for most concentration comparisons was low. (Source 11)

• Systematic review, Low certainty.
• Size: 81 trials in the network meta-analysis; 2 studies for the 2400-2800 ppm comparison.
• Who: children and adolescents, permanent dentition.
• How long: mostly 36 months.
• Result: 1700-2200 ppm versus 1450-1500 ppm SMD 0.04 (95% CI -0.07 to 0.15, indirect evidence only); 2400-2800 ppm versus 1450-1500 ppm SMD -0.05 (95% CI -0.14 to 0.05, two studies)
• Funding: not stated in the abstract we read.

For many comparisons of different concentrations the caries-preventive effects and our confidence in these effect estimates are uncertain and could be challenged by further research. The choice of fluoride toothpaste concentration for young children should be balanced against the risk of fluorosis.

The same review rated its own evidence weak and found almost nothing on harm from supplements. (Source 3)

• Systematic review, Very low certainty.
• Size: 11 studies, 7,196 children.
• Who: children under 16.
• How long: at least 2 years.
• Result: 10 trials at unclear risk of bias, one at high risk; effect on deciduous teeth unclear; limited information on adverse effects.
• Funding: not stated in the abstract we read.

We rated 10 trials as being at unclear risk of bias and one at high risk of bias, and therefore the trials provide weak evidence about the efficacy of fluoride supplements.

On the harms beyond teeth - skeletal fluorosis, fractures, bone maturity - the same Cochrane review could not reach any conclusion at all. (Source 10)

• Systematic review, Very low certainty.
• Size: 5 studies with incomplete participant numbers.
• Who: populations exposed to differing water fluoride concentrations.
• How long: not stated.
• Result: no usable estimate; the review states it was unsure of these effects.
• Funding: not stated in the abstract we read.

Because of very low-certainty evidence, we were unsure of other adverse effects (including skeletal fluorosis, bone fractures and skeletal maturity; 5 studies, incomplete participant numbers).

The same monograph was explicit about what it could not show: the animal and mechanistic evidence gave no clarity, the adult evidence was low confidence, and the effect of lower exposures is unresolved. (Source 12)

• Systematic review, Low certainty.
• Size: two high-quality cross-sectional studies in adults; heterogeneous and limited human mechanistic studies.
• Who: adults, and experimental animals.
• How long: not applicable.
• Result: low confidence for adult cognition; no determination possible on biological plausibility.
• Funding: US National Toxicology Program.

Existing animal studies provide little insight into the question of whether fluoride exposure affects IQ.

Most of the studies in that meta-analysis were judged at high risk of bias, and the authors said the dose-response below 1.5 mg/L is uncertain. (Source 14)

• Meta-analysis, Low certainty.
• Size: 74 studies, 52 rated high risk of bias and 22 low risk of bias.
• Who: children.
• How long: varied.
• Result: the authors state limited data and uncertainty below 1.5 mg/L when exposure is estimated from drinking water alone.
• Funding: not stated in the abstract we read.

There were limited data and uncertainty in the dose-response association between fluoride exposure and children's IQ when fluoride exposure was estimated by drinking water alone at concentrations less than 1.5 mg/L.

A 38-year prospective birth cohort in New Zealand found no difference in IQ between people who grew up with fluoridated water and those who did not. (Source 15)

• Cohort study, Moderate certainty.
• Size: a general population birth cohort born 1972-73 in Dunedin, with 95.4% retention at 38 years.
• Who: New Zealanders followed from birth.
• How long: 38 years; IQ measured repeatedly at ages 7 to 13 and at 38.
• Result: no clear IQ differences by fluoride exposure, holding after adjustment for sex, socioeconomic status, breastfeeding, birth weight and educational attainment. No effect size or interval is reported. The exposure studied is community water fluoridation at New Zealand levels together with fluoride toothpaste and 0.5 mg fluoride tablets, and the authors explicitly set their result against the separate literature on very high fluoride exposure, so this cohort is not evidence about high-fluoride exposure.
• Funding: not stated in the abstract we read.

These findings do not support the assertion that fluoride in the context of CWF programs is neurotoxic.

A 2026 meta-analysis restricted to community water fluoridation studies with adjusted regression coefficients found no association with IQ - but it pooled only three studies. (Source 16)

• Meta-analysis, Very low certainty.
• Size: 3 studies pooled out of 941 identified.
• Who: children (only one adult study existed, so no adult analysis)
• How long: observational.
• Result: pooled linear regression coefficient 1.01; heterogeneity between studies low.
• Funding: not stated in the abstract we read.

Limit of this finding: This result cannot carry the word 'safety'. The pooled figure is a linear regression coefficient reported as 1.01 with no confidence interval and no units given, so there is no way to tell how precise it is or what a one-unit change means. It pools three observational studies out of 941 screened, and there were too few adult studies to analyse at all. What it shows is that these three studies, pooled, found no association between community water fluoridation and IQ - which is an absence of evidence from very little evidence, not evidence that fluoridation is safe for IQ.

In the joint analysis, there was no association between exposure to fluoride in the water supply and IQ, for a linear regression coefficient (called β) of 1.01.

On cancer, the label's own summary of the epidemiology is that there is no credible evidence of a link, while rodent studies were equivocal in male rats only. (Source 17)

• Official position, Certainty not rated.
• Size: rodent carcinogenicity studies plus epidemiological data, numbers not given.
• Who: mice and rats; human populations with natural or added water fluoride.
• How long: lifetime rodent studies.
• Result: no carcinogenesis in mice of either sex or in female rats at 4.1 to 9.1 mg/kg; equivocal evidence in male rats at 2.5 and 4.1 mg/kg; a second study found none up to 11.3 mg/kg.
• Funding: manufacturer's own label (Westminster Pharmaceuticals)

Epidemiological data provide no credible evidence for an association between fluoride, either naturally occurring or added to drinking water, and risk of human cancer.

A 2026 review of fluoride's mechanism concludes that at the concentrations used in dental products there is no sign of genetic damage or harm to the oral microbiome. (Source 1)

• Expert review, not systematic, Low certainty.
• Size: not stated; a narrative synthesis of mechanism studies.
• Who: humans using standard topical fluoride products.
• How long: not applicable.
• Result: no genotoxic or adverse microbiome effects identified at recommended preventive concentrations; fluoride shifts the oral microbiome toward a health-associated state without reducing diversity.
• Funding: not stated in the abstract we read.

Despite well-known risks associated with high systemic intake, such as fluorosis, current evidence does not indicate genotoxic or adverse microbiome effects in humans from routine topical use of standard fluoride products at recommended preventive concentrations.

Where the evidence is mixed

Modern evidence on adding fluoride to drinking water shows a much smaller benefit than the pre-1975 studies did - roughly a quarter of one tooth - on low-certainty evidence. (Source 4)

• Systematic review, Low certainty.
• Size: 157 studies in the review, all non-randomised; 21 studies of fluoridation initiation; 2 studies and 2,908 children for the contemporary dmft estimate.
• Who: populations of all ages; the contemporary caries estimates are in children.
• How long: measured within three years of a change in fluoridation status and at end of follow-up.
• Result: change in dmft mean difference 0.24 (95% CI -0.03 to 0.52; P = 0.09), about one quarter of a tooth; proportion caries-free, 4 percentage points for primary and 3 percentage points for permanent dentition, both confidence intervals crossing no effect.
• Funding: not stated in the abstract we read.

This equates to a difference in dmft of approximately one-quarter of a tooth in favour of CWF; this effect estimate includes the possibility of benefit and no benefit.

The accompanying meta-analysis of 74 studies found an inverse association between fluoride and children's IQ overall, but the association for drinking water alone was null below 1.5 mg/L. (Source 14)

• Meta-analysis, Low certainty.
• Size: 74 studies (64 cross-sectional, 10 cohort); 59 studies and 20,932 children in the main pooled analysis; 20 studies for urinary fluoride.
• Who: children, mostly in China (45 studies), with studies from Canada, Denmark, India, Iran, Mexico, New Zealand, Pakistan, Spain and Taiwan.
• How long: prenatal or postnatal exposure windows.
• Result: pooled SMD -0.45 (95% CI -0.57 to -0.33) across group-level measures; for drinking water SMD -0.15 (95% CI -0.20 to -0.11) but null below 1.5 mg/L; individual-level urinary fluoride, IQ fell 1.63 points (95% CI -2.33 to -0.93) per 1 mg/L, and 1.14 points (95% CI -1.68 to -0.61) in low risk-of-bias studies.
• Funding: not stated in the abstract we read; the authors are the NTP review team.

In 31 studies reporting fluoride measured in drinking water, a dose-response association was found between exposed and reference groups (SMD, -0.15; 95% CI, -0.20 to -0.11; P < .001), and associations remained inverse when exposed groups were restricted to less than 4 mg/L and less than 2 mg/L; however, the association was null at less than 1.5 mg/L.

A 2026 systematic review of chronic fluoride exposure in young people splits the picture by dose: protective against decay at low concentrations, associated with harm at high ones. (Source 18)

• Systematic review, Low certainty.
• Size: 34 studies selected from 1,020 articles.
• Who: people aged 0 to 30.
• How long: chronic exposure.
• Result: low concentrations of 0.1-1.5 mg/L effective against caries; prolonged high-dose exposure associated with dental fluorosis, reproductive disturbances and cognitive or neurodevelopmental impairment; bone and endocrine findings inconsistent.
• Funding: not stated in the abstract we read.

Findings showed that while low fluoride concentrations (0.1-1.5 mg/L) effectively prevent dental caries, prolonged high-dose exposure is associated with adverse outcomes.

Where the research disagrees

Whether fluoride exposure at or near the levels used in water fluoridation lowers children's IQ

• US National Toxicology Program (2024 monograph), systematic review with confidence ratings of 72 human studies of fluoride and children's IQ; 19 were rated high quality and 18 of those reported an inverse association, and 46 of the 53 low-quality studies did too: This review finds, with moderate confidence, that higher estimated fluoride exposures (e.g., as in approximations of exposure such as drinking water fluoride concentrations that exceed the World Health Organization Guidelines for Drinking-water Quality of 1.5 mg/L of fluoride) are consistently associated with lower IQ in children. (Source 12)
• Taylor and colleagues (2025), the NTP team's own meta-analysis, on its own limits, meta-analysis of 74 studies, 64 cross-sectional and 10 cohort, mostly from China, 52 of them rated high risk of bias; in the drinking-water analysis the association was null below 1.5 mg/L: There were limited data and uncertainty in the dose-response association between fluoride exposure and children's IQ when fluoride exposure was estimated by drinking water alone at concentrations less than 1.5 mg/L. (Source 14)
• Broadbent and colleagues, Dunedin birth cohort (2015), prospective birth cohort followed 38 years with 95.4% retention, in a country with community water fluoridation; the exposure is community water fluoridation at New Zealand levels, fluoride toothpaste and 0.5 mg fluoride tablets assessed before age 5, and the authors explicitly distinguish their null result from the separate studies of very high fluoride exposure, so this side speaks only to fluoridation-level exposure: Results. No clear differences in IQ because of fluoride exposure were noted. These findings held after adjusting for potential confounding variables, including sex, socioeconomic status, breastfeeding, and birth weight (as well as educational attainment for adult IQ outcomes). Conclusions. These findings do not support the assertion that fluoride in the context of CWF programs is neurotoxic. Associations between very high fluoride exposure and low IQ reported in previous studies may have been affected by confounding, particularly by urban or rural status. (Source 15)
• Authors of a 2026 meta-analysis restricted to community water fluoridation studies, meta-analysis of observational studies; only 3 of 941 screened studies could be pooled, the pooled coefficient is reported with no confidence interval and no units, and no adult analysis was possible: In the joint analysis, there was no association between exposure to fluoride in the water supply and IQ, for a linear regression coefficient (called β) of 1.01. Heterogeneity between studies was low. This meta-analysis underlines the safety of community water fluoridation concerning IQ's intelligence assessment. (Source 16)

How much

• Reference intake: Dosing of a prescription fluoride product is set by the dentist or doctor. The label for 1.1% sodium fluoride gel (revised 08/2026) states it is for once-daily self-applied topical use, a thin ribbon on a toothbrush or in a mouth tray for at least one minute, preferably at bedtime, after brushing with ordinary toothpaste. That thin ribbon contains 2 mg of fluoride. This is the label's position. (Source 8)
• Upper limit: There is no drug upper limit on the label; the relevant ceiling is an environmental one. The World Health Organization's drinking-water guideline value is 1.5 mg/L of fluoride, and the US National Toxicology Program used exceeding that guideline as the exposure band at which it found a consistent association with lower IQ in children. That is a position of those bodies, with a date, not a dose for any reader. (Source 12)
• Studied: Toothpaste trials used concentrations from 440 ppm up to 2800 ppm fluoride; the dose-response for permanent teeth was demonstrated between 1000-1250 ppm and 2400-2800 ppm, with no further benefit above about 1500 ppm. (Source 2)
• Studied: Water fluoridation studies compared populations receiving fluoridated water with those receiving non-fluoridated or naturally low-fluoride water; the fluorosis estimates are anchored at a water fluoride level of 0.7 parts per million. (Source 10)
• Studied: The osteoporosis trial gave sodium fluoride 75 mg per day for four years, with a calcium supplement of 1,500 mg per day in both arms. (Source 5)

A common belief, and what the research shows

The belief: That because fluoride strengthens teeth it must strengthen bone, so more fluoride means stronger bones.

What the research shows: It was tested and the opposite happened. In a four-year randomised trial, sodium fluoride 75 mg a day raised lumbar spine bone density by 35% and yet "The number of new vertebral fractures was similar in the treatment and placebo groups (163 and 136, respectively; P not significant), but the number of nonvertebral fractures was higher in the treatment group (72 vs. 24; P less than 0.01)." The authors concluded that fluoride "increases cancellous but decreases cortical bone mineral density and increases skeletal fragility". A 2025 dose-response meta-analysis of water fluoride and fractures points the same way above about 1.5 mg/L.

Questions and answers

What is it?

Sodium fluoride is a simple inorganic salt, NaF - sodium plus fluoride ion. It is the fluoride compound in most toothpastes, in mouthrinses, in prescription high-strength dental gels and in fluoride tablets and drops, and it is one of the chemicals used to fluoridate drinking water. A prescription 1.1% sodium fluoride gel delivers 5 mg of fluoride ion per gram, and one thin ribbon of it contains about 2 mg of fluoride. (Source 19)

What does it do in the body?

Fluoride works mostly on the outside of the tooth rather than from within the body. Sitting in the saliva and plaque at low concentration, it slows the dissolving of enamel by acid and speeds its repair, and it forms calcium-fluoride-like reservoirs that release fluoride when the mouth turns acid. It also interferes with the enzymes and proton pumps of the bacteria that cause decay. High-concentration products increase enamel's resistance to acid and drive fluoride into the enamel surface. (Source 1)

Is it good or bad for you?

Both, and the dividing line is dose. At the low concentrations used in toothpaste and water it prevents decay: the Cochrane toothpaste review rates that high to moderate certainty. Above about 1.5 mg/L in drinking water, the evidence turns the other way: the US National Toxicology Program found with moderate confidence that higher exposures are consistently associated with lower IQ in children, and a dose-response meta-analysis found rising fracture risk above the same threshold. Even at 0.7 ppm, about 12% of people have dental fluorosis of cosmetic concern. At 75 mg a day it caused fractures. (Source 18)

How do you get more of it?

The studied sources are fluoride toothpaste, mouthrinse, professionally applied varnish and gel, prescription high-strength gel, swallowed tablets and drops, fluoridated drinking water, and tea. In trials, toothpaste at 1000 to 1500 ppm gave a dose-response benefit in children, while 5000 ppm products were used for high-risk situations such as root decay and dry mouth. None of this is a recommendation to any reader - the choice of concentration is a dental decision and has to be weighed against fluorosis risk. (Source 1)

If it is harmful, what reduces it?

For an acute swallow, calcium binds fluoride in the gut: the label says to give calcium such as milk for a small ingestion, and soluble calcium plus urgent medical help above 5 mg fluoride per kg of body weight. For chronic excess, the only approach is to cut the sources - defluoridated or alternative water, less tea, no supplements. Enamel defects already laid down by fluorosis are not something a treatment removes; they are a developmental change in the tooth. (Source 6)

Why might someone be low in it or missing it?

Someone gets little fluoride if their drinking water is naturally low in it and not fluoridated, if they use a non-fluoride toothpaste, or if they brush rarely. Because the main action is topical and repeated, intermittent exposure is effectively low exposure: the mechanism review stresses that the benefit depends on frequent, low-level exposure rather than on body stores. (Source 1)

Which whole foods contain it or feed it?

Drinking water is the main dietary source where supplies are fluoridated or naturally fluoride-rich, and the Cochrane water-fluoridation estimates in this write-up are all built on water concentration. Tea is the other food that comes up, but the evidence we found for it is plant science: a 2026 study of tea plants grown in solution describes Camellia sinensis as a hyperaccumulator of fluoride, meaning the leaf concentrates it, and the authors note the risk to humans from excess. That study measured plants, not people, and we found no human study quantifying how much fluoride tea drinking delivers, so no intake figure is given here. No whole food is a controlled source of fluoride in the way a toothpaste is. (Source 9)

What happens if you do not have it?

Without fluoride exposure, tooth decay is more common. The Cochrane review of 96 randomised trials put the difference in young children's primary teeth at 1.86 decayed or filled surfaces over about three years between fluoride toothpaste and non-fluoride toothpaste. Fluoride is not an essential nutrient in the sense that a deficiency disease exists - no syndrome of fluoride deficiency has been defined - so the consequence of having little is more decay, not an illness. (Source 2)

How can you test for it?

There are two practical measures. Urinary fluoride is the biomarker of total recent exposure and is what the neurodevelopment studies used; in the 2025 meta-analysis of individual-level data, IQ fell 1.63 points per 1 mg/L rise in urinary fluoride. The other is looking at the teeth: dental fluorosis is scored clinically and is itself used as a marker of past exposure. Neither tells you body stores, and urinary fluoride varies with what you drank that day, so a single reading is a weak measure of long-term exposure. (Source 14)

References

1. Dentistry journal. The Cariostatic Mechanisms of Fluoride-An Updated Review.. 2026. PMID 42505698, DOI 10.3390/dj14070390. Read the source
2. The Cochrane database of systematic reviews. Fluoride toothpastes of different concentrations for preventing dental caries.. 2019. PMID 30829399, DOI 10.1002/14651858.cd007868.pub3. Read the source
3. The Cochrane database of systematic reviews. Fluoride supplements (tablets, drops, lozenges or chewing gums) for preventing dental caries in children.. 2011. PMID 22161414, DOI 10.1002/14651858.cd007592.pub2. Read the source
4. The Cochrane database of systematic reviews. Water fluoridation for the prevention of dental caries.. 2024. PMID 39362658, DOI 10.1002/14651858.cd010856.pub3. Read the source
5. The New England journal of medicine. Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis.. 1990. PMID 2407957, DOI 10.1056/nejm199003223221203. Read the source
6. DailyMed / Westminster Pharmaceuticals, LLC (US prescribing information, revised 08/2026). SODIUM FLUORIDE 1.1% GEL (sodium fluoride) gel - FDA prescribing information (Adverse Reactions and Overdosage). 2026. Read the source
7. DailyMed / Westminster Pharmaceuticals, LLC (US prescribing information, revised 08/2026). SODIUM FLUORIDE 1.1% GEL (sodium fluoride) gel - FDA prescribing information (Contraindications and Warnings). 2026. Read the source
8. DailyMed / Westminster Pharmaceuticals, LLC (US prescribing information, revised 08/2026). SODIUM FLUORIDE 1.1% GEL (sodium fluoride) gel - FDA prescribing information (Overdose treatment dose and Dosage and Administration). 2026. Read the source
9. Plant physiology. Fluoride detoxification in tea plants depends on aluminum and localization in the epidermis.. 2026. PMID 41722032, DOI 10.1093/plphys/kiag077. Read the source
10. The Cochrane database of systematic reviews. Water fluoridation for the prevention of dental caries. [fluorosis, other adverse effects and cessation]. 2024. PMID 39362658, DOI 10.1002/14651858.cd010856.pub3. Read the source
11. The Cochrane database of systematic reviews. Fluoride toothpastes of different concentrations for preventing dental caries. [background, adverse effects and conclusions]. 2019. PMID 30829399, DOI 10.1002/14651858.cd007868.pub3. Read the source
12. NTP monograph. NTP monograph on the state of the science concerning fluoride exposure and neurodevelopment and cognition: a systematic review.. 2024. PMID 39172715, DOI 10.22427/ntp-mgraph-8. Read the source
13. Environmental health : a global access science source. The association of fluoride exposure with bone density and fracture risk: a dose-response meta-analysis.. 2025. PMID 41068882, DOI 10.1186/s12940-025-01226-y. Read the source
14. JAMA pediatrics. Fluoride Exposure and Children's IQ Scores: A Systematic Review and Meta-Analysis.. 2025. PMID 39761023, DOI 10.1001/jamapediatrics.2024.5542. Read the source
15. American journal of public health. Community Water Fluoridation and Intelligence: Prospective Study in New Zealand.. 2015. PMID 24832151, DOI 10.2105/ajph.2013.301857. Read the source
16. Ciencia & saude coletiva. Community water fluoridation and intelligence quotient: systematic review and meta-analysis of observational studies.. 2026. PMID 41779589, DOI 10.1590/1413-81232026312.19682023. Read the source
17. DailyMed / Westminster Pharmaceuticals, LLC (US prescribing information, revised 08/2026). SODIUM FLUORIDE 1.1% GEL (sodium fluoride) gel - FDA prescribing information (Precautions: carcinogenesis, mutagenesis, pregnancy). 2026. Read the source
18. Journal of environmental science and health. Part C, Toxicology and carcinogenesis. Health implications of chronic fluoride exposure in children and young adults: a systematic review.. 2026. PMID 42418311, DOI 10.1080/26896583.2026.2697147. Read the source
19. DailyMed / Westminster Pharmaceuticals, LLC (US prescribing information, revised 08/2026). SODIUM FLUORIDE 1.1% GEL (sodium fluoride) gel - FDA prescribing information (Description, Clinical Pharmacology, Indications, Contraindications and Warnings). 2026. Read the source