Nano-hydroxyapatite and remineralisation: what happens to softened enamel.

Softened enamel — enamel that acid has pulled mineral out of without yet wearing it away — can take that mineral back, and nano-hydroxyapatite is one of the materials that can supply it. No toothpaste rebuilds enamel: enamel has no cells to regrow it, so what goes back in is mineral, into a surface that is still there. Almost every result below is a hardness reading or a mineral proxy taken on extracted or bovine teeth over days or weeks. The independent syntheses and the manufacturer-linked ones disagree about how much hydroxyapatite adds over fluoride, and the disagreement runs along the funding line rather than through the numbers. On acid erosion the answer is plainer and worse: a 2026 umbrella review of eight systematic reviews rated the certainty of the evidence for calcium-based and biomimetic technologies low to very low, against moderate for stannous formulations¹. In S3 Sensitivity Science™ the fluoride stays in at 1450 ppm and takes the monofluorophosphate form, beside two hydroxyapatites.

What was checked24 peer-reviewed studies, Oral Health Foundation and NHS guidance

Can a toothpaste remineralise softened enamel?

Within a narrow meaning of the word, yes. The Oral Health Foundation sets out the cycle the word belongs to: every time you eat or drink something acidic the enamel softens for a while and loses some of its minerals, saliva usually repairs that damage, and the trouble starts when the acid arrives more often than the surface can recover from⁵. A remineralising toothpaste is aimed at the recovery half of that cycle. It supplies calcium and phosphate, or something that helps the surface hold on to them, while the softened layer is still in place.

That is a smaller promise than the word "repair" carries in a shop. Enamel is not living tissue and has no cells left to grow it back once a tooth has come through. What every paper on this page measures is mineral going into a surface, or a lesion not getting worse, and never a tooth restored. The Journal's explainer on tooth enamel erosion covers what happens when the loss goes beyond softening.

Even in the laboratory, recovery is partial. In an in-vitro study on 60 bovine incisors given an artificial caries challenge, two weeks of daily treatment during pH cycling with six preparations — a hydroxyapatite toothpaste and a hydroxyapatite-with-fluoride toothpaste among them — raised surface hardness by between 18.7% and 35.2%, and not one of them brought the enamel back to where it started⁶. That study also found the same elemental composition on the treated surfaces as on the untreated control; only the calcium and phosphorus depth profiles separated them⁶.

What happens when hydroxyapatite meets a softened surface?

It behaves as a mineral donor and as a filler. The crystal is the same compound the enamel is built from, so particles small enough to enter the porosity of a softened layer can sit there and give up calcium and phosphate ions locally; the larger, less crystalline biomimetic form works instead as a layer laid across the top. Why the particles stick at all is a question about crystal shape and lattice match, and it is answered on the page about rod-shaped crystals and the enamel lattice.

Two in-vitro studies map the size of the effect on softened enamel. Brushed twice a day for two and five weeks in a remineralising solution, toothpastes carrying nano-hydroxyapatite remineralised bovine dentine lesions more than an amine fluoride toothpaste, and a 7% pure nano-hydroxyapatite paste did the same for enamel⁷. The most instructive arm of that study had no toothpaste in it: specimens left in the mineral solution alone gained as much as those brushed with the nano-hydroxyapatite pastes⁷. In vitro, in a ten-day pH-cycling model on 60 extracted human incisors, a nano-hydroxyapatite dentifrice recovered more surface microhardness than bioactive glass, casein phosphopeptide-amorphous calcium phosphate and fluoride dentifrices⁸ — though the paper publishes the order of the four and not the numbers behind it, which is worth knowing before the ranking is quoted anywhere.

Move from a beaker to a mouth and the differences shrink. In a randomised double-blind in-situ crossover study, 30 adults wore appliances carrying demineralised enamel blocks for 28 days per phase; a 5% nano-hydroxyapatite dentifrice, a 10% one and an 1100 ppm fluoride dentifrice all reduced mineral loss and lesion depth significantly, and analysis of variance found no significant difference in mineral gain among the three⁹.

What has been measured, and how?

Mostly hardness and mineral proxies, over short periods, on teeth that are not in anyone's head. A 2026 scoping review of 122 studies of fluoride-free toothpaste alternatives put the state of the field in one line: the evidence is dominated by surrogate outcomes and short-term study designs, with hydroxyapatite carrying the most consistent evidence base of the fluoride-free categories and none of them matching fluoride for long-term caries prevention in children¹⁰. The table below is the whole evidence base this page rests on, with the column that usually goes missing — what was actually measured — and the column that decides how much weight a result carries.

Two things are worth taking from that table. Almost every row measures a proxy — hardness, mineral loss, a fluorescence reading, a lesion count — and the three rows that follow people for a year or more are all counting caries lesions rather than watching the softened surface this page is named after. The bench results also refuse to line up the way a shopper would expect: in an industry-funded laboratory comparison of eight marketed toothpastes on 160 bovine enamel blocks over 14 days, whose corresponding author reports financial support from Oral Microbiome Solutions, every product raised surface microhardness significantly and the order did not follow the percentage printed on the tube¹¹.

Where do the independent studies disagree with the manufacturer-linked ones?

They disagree about the conclusion, not about the numbers, and the split runs almost exactly along the funding line.

The manufacturer-linked syntheses read the evidence as settled enough to act on. A 2024 systematic review and meta-analysis pooled five clinical and eight in-situ trials and concluded that hydroxyapatite works as an anti-caries active in the absence of fluoride; two of its authors are scientists employed by Dr Wolff, a maker of hydroxyapatite toothpastes, and every author on it reports travel grants from that company¹². A 2021 systematic review and meta-analysis from the same group, funded by that manufacturer, pooled three randomised trials and reported 17% caries protection¹³. Two randomised trials sit underneath them. The first is an 18-month double-blind trial in 189 adults, with manufacturer-affiliated authors, in which 89.3% of the hydroxyapatite group and 87.4% of the fluoride group finished with no increase in decayed, missing and filled surfaces¹⁴. The second was a randomised trial that ran 168 days in young people wearing fixed braces, paid for by the maker of the dentifrice it tested, and it found no significant difference from a 1400 ppm fluoride paste¹⁵.

The independent work is flatter. The 2025 meta-analysis of four randomised trials found no difference in new and progressing lesions and a pooled risk ratio of 0.98³. The 2022 systematic review could not pool its in-vivo studies at all, rated the level of evidence very low, and reported an in-situ meta-analysis in which nano-hydroxyapatite lost significantly more mineral than sodium fluoride under demineralising conditions (mean difference 1625, 95% CI 553 to 2697) while matching it under remineralising ones (−15, 95% CI −133 to 103)². Its authors put their own finding this way: nano-hydroxyapatite may not hinder demineralisation, but it may be a way of remineralising enamel where remineralising conditions already exist². In a 14-day laboratory model on 90 bovine enamel specimens, a fluoride-free nano-hydroxyapatite toothpaste was the one paste of five that failed to reduce demineralisation against a fluoride-free control, and the specimens discarded for surface loss came mainly from that group and the control¹⁶.

Not every independent result is unfriendly. An independent commentary in Evidence-Based Dentistry in 2025 read the manufacturer-linked review and reported its pooled clinical comparison with fluoride toothpaste as favouring hydroxyapatite without reaching significance, at an odds ratio of 1.1, and concluded that hydroxyapatite toothpastes can be considered a good alternative for caries¹⁷. And a randomised trial in 92 adults with 546 initial approximal lesions, with no conflict declared, found a home nano-hydroxyapatite gel remineralising 36.5% of those lesions on radiographs at one year — with the caution that the gains in its combined ozone arm had reversed by two years¹⁸.

Read across the two columns, one pattern holds up. Where hydroxyapatite is tested instead of fluoride, the independent answer is "no detectable difference" and the manufacturer-linked answer is "it works". Where it is tested alongside fluoride, the independent answer is warmer. A 2024 systematic review of five studies and 151 white spot lesions found moderate evidence favouring nano-hydroxyapatite combined with fluoride, while its authors declined to recommend extended use on what they had¹⁹. The triple-blind trial that followed 610 children for 24 months set a hydroxyapatite-fluoride toothpaste against a monofluorophosphate one at the same fluoride level, and by the end the hydroxyapatite-fluoride group had significantly fewer enamel lesions⁴. That is a real pattern rather than an average of the two, and it carries a caveat: it is also the shape of the formula S3 sells, which is why it sits here as a reading of the evidence and not as proof of anything about a product.

Does it protect against acid erosion?

No, on the evidence there is, and this is the part of the subject where hydroxyapatite does worst.

Erosion is not the same event as caries. A caries lesion is demineralised from below a surface that stays intact, which is what leaves something for mineral to go back into; erosive wear takes the surface itself away, and the umbrella review of eight systematic reviews describes it as irreversible¹. That review rated the certainty of the evidence moderate for stannous-containing formulations and low or very low for the calcium-based and biomimetic technologies, hydroxyapatite among them, and warned that the experimental findings should not be read as proof of reduced erosive wear in patients¹.

The two laboratory studies underneath that rating are blunter. In vitro, in a ten-day erosion and abrasion model on human enamel with citric acid six times a day, none of four hydroxyapatite toothpastes and a hydroxyapatite mouthrinse reduced tissue loss against an active-free control toothpaste, two of them increased it significantly (11.7 and 13.0 µm against 7.8 µm), and a fluoride and stannous mouthrinse cut it to 1.3 µm; brushing raised tissue loss in every group except that rinse²⁰. Its authors conclude that for erosive tooth wear, hydroxyapatite looks like neither a replacement for fluoride nor a useful addition to it²⁰. An earlier model of sixteen toothpastes reached a compatible place: most reduced tissue loss when applied as slurries, brushing raised it in almost all of them, and only the three stannous products and two sodium fluoride toothpastes still reduced loss when the specimens were brushed²¹.

Both are laboratory models with a severe acid challenge, and neither tells you what happens to a person's teeth over a year. What they do rule out is the marketing shape of the idea — that a mineral toothpaste puts down a sacrificial layer that takes the acid instead of the enamel. The same in-vitro analysis that measured the tissue loss also looked at what is in the tubes: the particulate fraction of every toothpaste examined was dominated by silicon and oxygen, with calcium and phosphorus present only in small amounts on the particles²⁰. For erosion, the guidance-level advice has not changed, and the Oral Health Foundation's version of it is behavioural rather than chemical: keep acidic food and drink to mealtimes, and wait at least an hour before brushing so the surface can harden again⁵.

Why does S3 keep fluoride in a hydroxyapatite toothpaste?

Because the independent evidence is friendliest to the combination, and because the decay evidence has never moved. The National Health Service's advice for looking after teeth is still to brush twice a day with a fluoride toothpaste²², and nothing in the remineralisation literature displaces that: the 2026 scoping review's conclusion was that no fluoride-free category has evidence comparable to fluoride for long-term caries prevention in children¹⁰.

S3 combines a nerve-calming active, 5% potassium nitrate, with nano-hydroxyapatite at 10% and biomimetic hydroxyapatite at 5%, both as solution, and full adult-strength fluoride in the same tube. The fluoride is 1450 ppm, delivered as sodium monofluorophosphate. That salt is not an accident of supply: monofluorophosphate does not react with the calcium in hydroxyapatite inside the tube the way a free fluoride ion would. The two hydroxyapatites are sized for two different jobs, one under 100 nm for the tubule and one around two microns and less crystalline for the surface. A fluoride-free hydroxyapatite paste asks you to trade decay protection for mineral, and this formula does not.

None of that is a claim that the combination outperforms fluoride alone. Nobody has run that trial on this formula. The whole argument about hydroxyapatite set against fluoride, rather than beside it, belongs on the page comparing the two ingredients directly, and the case for carrying both is set out on the page on toothpastes that contain fluoride and hydroxyapatite together.

Does remineralising enamel reduce sensitivity?

No, or at least nobody has shown that it does. Mineral going back into enamel and a person's sensitivity easing have never been watched in the same mouths and tied to each other.

The two bodies of evidence share an ingredient and almost nothing else. One of them measures hardness, mineral loss and lesion counts on enamel. The other measures what a person reports when cold air or a probe reaches exposed dentine. An independent 2019 meta-analysis of six randomised trials read at four weeks put nano-hydroxyapatite ahead of its comparators on air and touch, and level with them on cold²³. A larger 2023 meta-analysis, whose author list includes employees of a hydroxyapatite toothpaste manufacturer, pooled 44 clinical trials and reported a 39.5% fall in sensitivity against placebo²⁴. Neither took a mineral measurement of any kind. The weighing of that sensitivity evidence, funding column and all, is on the page asking whether hydroxyapatite toothpaste works for sensitive teeth.

The one trial that set out to do both is worth reading closely, because its title promises more than its methods deliver. Forty adults with white spot lesions and air-evoked sensitivity on those teeth were randomised to a biomimetic hydroxyapatite toothpaste or a 1450 ppm fluoride toothpaste for 90 days; the Schiff air index and the visual analogue score fell significantly in the hydroxyapatite group and not in the fluoride group, and an erosive wear index did not shift in either²⁵. Two things it did not do: it reported no mineral, mineral-loss or lesion-depth measurement of any kind, and the paper states that no significant difference was found between the two groups on either sensitivity measure at any time point²⁵. The within-group time course is what changed, and the between-group comparison is what a reader wants.

Where the two subjects genuinely touch is the tubule rather than the enamel surface: mineral that settles inside an open dentine tubule slows the fluid movement a cold drink sets off, which is a mechanism the page on how nano-hydroxyapatite seals dentine tubules takes apart step by step. Hardening a softened patch of enamel on the outside of a tooth is a different event, and it has never been shown to change what the nerve underneath does.

What can remineralisation not do?

It cannot put back what is no longer there, and it cannot work on a surface that is already gone. Those two limits are the whole of the practical answer, and this is where they land.

Two practical consequences follow. Softening is a daily event and the recovery window matters more than the paste: keep acids to mealtimes and leave an hour before brushing, so the surface has hardened before a brush reaches it⁵. And a tooth that has become sensitive quickly, or that looks visibly different, is a reason to see a dentist rather than to change toothpaste; the NHS is direct about what untreated problems cost, saying that damage left long enough becomes harder or impossible to repair²². How long a mineral effect takes to show up at all is a separate question, answered on the page on how long nano-hydroxyapatite toothpaste takes to work.

Frequently asked questions

Does S3 rebuild enamel?

No, and neither does any other toothpaste. Enamel has no cells to regrow it, so nothing in a tube restores tissue that has been eroded or abraded away. What hydroxyapatite can do is supply mineral to a surface that has softened and kept its structure, and what has been measured of that is hardness recovery and mineral uptake, mostly on extracted or bovine teeth over days or weeks, with partial recovery even in the laboratory⁶. S3 states what is in the tube — the fluoride level, the two hydroxyapatites, the potassium nitrate — and not an enamel outcome.

Is hydroxyapatite better than fluoride for remineralisation?

On caries outcomes the pooled randomised evidence cannot tell the two apart, at a risk ratio of 0.98 with a confidence interval running from 0.85 to 1.12³. The laboratory evidence points both ways depending on the model, which is why the table above prints what each paper measured rather than a verdict. On erosion specifically, fluoride and stannous formulations are ahead and hydroxyapatite has not protected enamel in the models that tested it¹²⁰. The full comparison is on the page setting the two ingredients against each other.

Can a toothpaste reverse a white spot?

It can sometimes reduce one, and it rarely removes the mark. Across five studies and 151 white spot lesions, a 2024 systematic review found moderate evidence for nano-hydroxyapatite combined with fluoride while stopping short of recommending extended use¹⁹. A separate 2024 meta-analysis, pooling 14 studies of which ten were laboratory work, found no improvement in the colour of the lesions at all (mean difference −2.76, 95% CI −6.79 to 1.27)²⁶. A white spot that has been there for years is unlikely to disappear from brushing.

Does remineralising toothpaste help with sensitivity?

Not by the same route, and nobody has tested the link. Relief in the sensitivity work comes from what happens at exposed dentine — tubules occluded, the nerve calmed — rather than from mineral going back into enamel on the outside of a tooth. The one trial that set out to cover remineralisation and sensitivity together took no mineral reading at all, and found no significant difference between its two arms on either sensitivity measure²⁵.