Potassium nitrate and hydroxyapatite in one tube: what the pooled estimate rests on, and what nobody has tested.

A toothpaste carrying potassium nitrate and hydroxyapatite together is aimed at two different parts of the same tooth — the potassium at the nerve endings inside the dentine tubule, the mineral at the open mouth of it — and the pooled figure that pairing is usually quoted on comes from one node of a 2020 systematic review and network meta-analysis of 125 randomised trials, a node built on two randomised trials and 140 patients¹. The estimate is real and it is thin: the same network meta-analysis gave it moderate certainty and marked it down for imprecision, and the newest and largest synthesis in the field, a 2026 systematic review and network meta-analysis of 93 randomised trials and 9,548 participants, carries no potassium-plus-hydroxyapatite category at all². Of 51 UK sensitivity toothpastes whose published ingredient lists were read for the category review in September 2026, S3 Sensitivity Science™ is the only one that carries potassium nitrate, hydroxyapatite and fluoride in one formula — a fact about ingredient lists, and not evidence that the combination relieves more than either active on its own.

What was checked16 peer-reviewed studies, the Oral Health Foundation, the NHS, and product information as published by each brand

Why would you put a nerve active and an occluder in the same tube?

Because the two mechanisms answer different halves of the problem, and one of those halves is demonstrated while the other is only proposed. A sensitive tooth has a nerve that fires too readily and dentine tubules that are open, and most sensitivity toothpastes take on one of those and leave the other. The page on why sensitivity is two problems rather than one argues that case at length, and the Journal's article on the two causes of sensitive teeth makes it for a general reader. This page counts what stands behind it.

The occlusion half is visible. Put a mineral paste on a dentine surface, brush it, and an electron microscope shows the tubules narrowed or filled. The nerve half is not visible, and it has never been watched happening. A narrative review of 27 clinical trials, sixteen of them double-blind randomised trials of potassium toothpastes, states plainly that the mechanism — potassium ions diffusing along the tubule until the nerve endings stop responding readily — is a hypothesis that has not been confirmed in an intact human tooth¹². It is also the working assumption behind every potassium product on the shelf, and the page on how potassium nitrate is thought to calm a tooth takes it apart properly.

The two do not overlap. In a four-week double-blind randomised trial of 30 adults with a companion electron-microscope test, a 5% potassium nitrate toothpaste reduced air-blast and cold-water sensitivity more than a non-desensitising paste while the microscope found the tubules still open¹³. So the case for one tube does not rest on arithmetic. It rests on the observation that a potassium salt leaves a hole unsealed and a mineral leaves a nerve untouched, and that a reader with both problems is being sold half an answer twice. That is a mechanistic argument. Whether it produces more relief in a mouth is a different question, and the rest of this page is about how little has been done to settle it.

What does the pooled evidence actually say about the combination?

It says the pairing looks strong and that almost nothing sits underneath it, and those two halves are usually quoted apart. The 2020 review is the source everyone reaches for. Across the 125 randomised trials and 12,541 patients in that network meta-analysis, each class of desensitising toothpaste was compared against a plain fluoride toothpaste, and potassium with hydroxyapatite came out near the top: a standardised mean difference of 2.47 on the touch test and 2.44 on the air blast, both with moderate certainty¹.

Opening the review's own supplementary appendix changes how that reads. The network meta-analysis records the node as two randomised trials and 140 patients on the touch test and the same two on the air blast, and it gives the reason for the downgrade as imprecision: the credible interval runs from 0.3 to 4.64 on touch and from 0.33 to 4.55 on air, so the lower end sits a hair above no effect at all¹. In the same tables the stannous fluoride estimate rests on 12 trials and 1,724 patients and the potassium estimate on 14 trials and 1,138 patients¹. Counted by treatment arm, potassium with hydroxyapatite is two of the 326 arms in the whole review, against 66 for potassium alone and five for hydroxyapatite alone¹. There is a third stimulus, cold, pooled across only 16 studies and 1,093 participants: the pairing does not appear there at all, because no trial supplied the data¹.

The appendix that lists all 125 included trials does not map trials to nodes, so this page cannot name the two¹. It can say what it looked for. Six of the included trials test a hydroxyapatite paste, and the abstracts of five of them were read for this page: they run a mineral paste against arginine, against strontium chloride, against a benchmark toothpaste, or against fluoride and a placebo, and none of the five puts a potassium salt in the test paste as well. The sixth was published in a Korean journal that carries no abstract in any index this project can reach, so it stays an open box rather than a guess.

The 2026 review is the check on all of it: 93 randomised trials, 9,548 participants, and a node list that runs stannous fluoride, arginine, nano-hydroxyapatite, bioglass, potassium with or without fluoride, potassium plus stannous fluoride, oxalate and strontium acetate, with nothing for the mineral-and-potassium pairing². Its own nodes are thin where the earlier one is thin: in that 2026 network meta-analysis nano-hydroxyapatite rests on two studies and potassium on one, against ten for stannous fluoride, the only high-confidence result in the whole network². Its authors put stannous fluoride and arginine first for self-care and say the choice between actives should follow preference, tolerability and availability rather than an expectation that one beats another². Neither of those two classes is in S3's formula. Between the two reviews lie six years in which the combination question was not asked again.

Two other syntheses belong beside them, because they disagree with each other and the disagreement is the finding. A 2019 network meta-analysis of 30 randomised trials ranked nano-hydroxyapatite likeliest to be the most effective single active at the two-week and four-week readings, while finding potassium toothpaste no different from placebo in the same network¹⁴. And a systematic scoping review that catalogued 368 toothpaste formulations across 138 randomised trials, counting formulations rather than pooling results, found 68 with a potassium compound, nine with hydroxyapatite, and nineteen — out of 368 — in which potassium sat alongside a second desensitising ingredient of any kind¹¹. That nineteen is the whole combination literature, for every pairing anyone has tried¹¹.

Which trials have put both actives in one paste?

Seven, on a generous reading, and each one has something that stops it settling the question. The table below is the whole set, and the last column is the point of it.

Three of those rows deserve a sentence each. Amaechi's is the most valuable thing on the page, because it changes one ingredient and holds the rest of the formula still: the same 10% nano-hydroxyapatite base, with and without the potassium, in randomised groups over eight weeks, in a trial funded by the company that made the pastes³. It answers the mineral-side question only. Nobody has run the mirror image, the pairing against the potassium alone, which is the arm that would tell a reader whether the mineral is earning its place.

Ishii's trial shows that the missing arm can be built, and it was the manufacturer that built it⁵. Scientists employed by that manufacturer set 5% potassium nitrate against 5% potassium nitrate plus a second active, with a placebo underneath both, and reported the two-active paste ahead at four and eight weeks⁵. The second active there is aluminium lactate. Read next to Seong's three-active paste⁴, it is also the reason Seong's large numbers cannot be credited to the hydroxyapatite: when the aluminium lactate was isolated in a trial of its own, it did work.

And Low's is the one that most resembles the formula this page is about, which is exactly why it has to be described accurately⁶. Nano-hydroxyapatite, potassium nitrate and monofluorophosphate in one marketed paste, self-reported improvement at 48 hours and again at a fortnight — and no control group, no randomisation, no blinding and no instrument reading⁶. An uncontrolled before-and-after design, in a condition whose placebo response has been measured and is large¹⁴, cannot tell you the formula did the work. It can tell you the formula exists and that somebody measured people using it.

One more trial belongs here, because it takes the two actives apart instead of putting them together. In a four-week non-blinded randomised trial of 39 participants, a hydroxyapatite toothpaste and a potassium nitrate toothpaste both relieved air-blast sensitivity, and the mineral paste came out comparable to the potassium one¹⁰. It is small and it was not blinded, so it can neither establish nor rule out a difference. It is still the most direct head-to-head anyone has published, and its answer is a shrug.

There is one further line of work, and it is not clinical. A potassium salt and hydroxyapatite have been engineered into a single material: in the laboratory, an experimental varnish combining potassium chloride with fluoridated hydroxyapatite released potassium over six hours and cut dentine permeability more than a commercial fluoride varnish, with tubules occluded on the microscope afterwards¹⁵. Gels pairing nano-hydroxyapatite with 5% potassium nitrate have been characterised in vitro on extracted human teeth, where they raised surface hardness and partly restored mineral content¹⁶. Both are mechanism, on dentine that was never attached to anyone, and neither says anything about relief.

Is a potassium salt plus hydroxyapatite the same thing whichever salt it is?

Not on the evidence, and the distinction matters here more than almost anywhere else: two of the three combination trials with a modern randomised design used potassium citrate rather than nitrate⁸⁹, the larger of them funded by a toothpaste manufacturer that employs four of its authors⁸. Potassium is thought to be the working ion, which is why the three salts are grouped together in guidance and in reviews. Nobody has put potassium nitrate and potassium citrate side by side in one randomised comparison, and the page on the two potassium salts sets out what each has and has not been tested at.

For this page the consequence is narrow and concrete. A reader who takes "potassium plus hydroxyapatite has a large pooled effect" from the 2020 network meta-analysis and then buys a citrate-and-hydroxyapatite paste has bought a formula that review did not pool¹. A reader who buys a nitrate-and-hydroxyapatite paste has bought one that review barely pooled, since the node holds two randomised trials¹. The honest reading is that the question was put once, to a very small pile of evidence, and has not been put again¹.

The Oral Health Foundation's page on sensitive teeth shows how easily the two mechanisms get flattened into one. It names potassium citrate, potassium nitrate and stannous fluoride as the actives that help, and then explains that sensitive toothpastes "work by blocking the tiny channels in dentine"¹⁷. That is accurate for stannous fluoride and for the minerals. It is not what a potassium salt does, and the electron-microscope work says so¹³. The guidance is not wrong about what to buy. It simply does not separate the two jobs, and that separation is what this page turns on.

Which UK toothpastes actually combine them, and what do they give up to do it?

Six of the 49 products in the September 2026 ingredient scan, and five of those six give something up to get there. The table below comes from that scan, based on what each brand states about its own formula.

Two things fall out of that table. The first is the salt: one row uses potassium nitrate, and the near miss is a UK brand on the same shelf whose potassium is citrate. That is why the register's wording says potassium nitrate and is never loosened to "a potassium salt". A seventh product, Pärla Pro, describes potassium citrate, hydroxyapatite and 1450 ppm fluoride in a tablet, but publishes no full ingredient list, so it stays off the table rather than being guessed at.

The second is fluoride. Four of the six rows carry none at all. That is a decision about decay, not about sensitivity: dropping fluoride costs a user the protection the NHS recommends fluoride toothpaste for, and changes nothing in the sensitivity literature either way. Unlike fluoride-free hydroxyapatite pastes, S3 does not ask you to give up decay protection. The chemistry behind keeping both is that monofluorophosphate does not react with the calcium in hydroxyapatite inside the tube, which is why the fluoride salt on a combination label is worth a glance. This page compares ingredients and stated actions only, not clinical performance. Prices and formulations may change; always check the pack.

What has and has not been tested about S3's own combination?

Nothing has been tested about it in people, and that is the sentence this page exists to make sure a reader gets from us before they get it from somebody else. There is no randomised controlled trial of the S3 formula, no comparison of it against either of its own actives, and no entry for it in either network meta-analysis, because a network meta-analysis pools trials and there are none of this formula to pool. The 2020 network meta-analysis estimate that suits the architecture so well was calculated from two randomised trials of other people's pastes¹.

Two pieces of evidence do exist, and they show different sorts of thing. The formula has been tested at the University of Reading, which is laboratory work on dentine: it can show what a paste does to a surface, in the same category as the varnish and the extracted-tooth work above¹⁵¹⁶, and it is not a measurement of anyone's comfort. And there is an independent third-party consumer trial of 51 adults with sensitive teeth over eight weeks, which records what those adults said about their own teeth. Neither of those is a controlled test of the pairing, and no percentage from either appears anywhere on this page, because a survey of users cannot settle a question about mechanism and because the argument here is built out of other people's controlled trials.

The declared levels are worth stating precisely, since they are the real basis of the mechanistic case. 10% nano-hydroxyapatite and 5% biomimetic hydroxyapatite are inclusion levels of the ingredient as supplied; the active hydroxyapatite content is lower, and S3 states both. Read against the literature, the formula sits at inclusion levels that match the published studies: 5% potassium nitrate, 10% nano-hydroxyapatite solution. That is a reasonable thing to build. It is not a result.

What should you look for on a label?

Four things, in this order, and none of them is a claim on the front of the box.

The first is which potassium salt it is. Nitrate, citrate and chloride are grouped as one class in guidance, have been tested separately, and have never been put against each other, so a pack that says only "potassium" is telling you less than it looks.

The second is whether there is a fluoride at all, and which one. A hydroxyapatite paste without fluoride is a decay decision as much as a sensitivity one.

The third is whether any level is declared. Most of the shelf declares nothing: on the table above, one row publishes levels for all three of its actives, one publishes a fluoride concentration and nothing else, and four publish none. A level you cannot read is a level you cannot set against the ones in the published literature.

The fourth is time. Potassium is judged at two to four weeks, and nothing in this literature supports a first-week verdict on it. If the pain is severe, if sensitivity lasts more than a few weeks, if only one tooth is affected or if the pain came on suddenly, the Oral Health Foundation says to see a dentist¹⁷; the NHS puts toothache lasting more than two days in the same place¹⁸. Those are the cases where nothing on this page applies.

Frequently asked questions

Is a toothpaste with both potassium nitrate and hydroxyapatite better than one with either on its own?

Nobody has shown that, and one randomised comparison has done half of it. In an eight-week double-blind randomised trial of 85 adults, funded by the maker of the pastes it tested, a 10% nano-hydroxyapatite toothpaste with 5% potassium nitrate added eased cold-stimulus sensitivity more than the identical paste without it at weeks two, four and six³. That is the pairing against the mineral alone. The pairing against the potassium alone has never been published, and the figure that looks like an answer is a network meta-analysis node resting on two randomised trials and 140 patients¹.

Has S3's own combination been tested in people?

Not in a controlled test, and no page of ours should be read as implying one. What exists is laboratory testing at the University of Reading, which measures what the paste does to a dentine surface, and an independent third-party consumer trial of 51 adults with sensitive teeth over eight weeks, which records what those adults said about their own teeth. Neither is a controlled comparison of the pairing, and this page quotes no figure from either.

Which UK toothpastes contain both a potassium salt and hydroxyapatite?

Six were found in the September 2026 ingredient scan of 51 products: S3, and five others whose potassium is citrate or chloride rather than nitrate. Four of those five carry no fluoride. A seventh, Pärla Pro, states the trio in a tablet but publishes no ingredient list, so it is recorded as unconfirmed.

Is potassium citrate with hydroxyapatite the same thing as potassium nitrate with hydroxyapatite?

Nobody knows, because the two salts have never been put in the same randomised comparison. Both have been tested separately, both have beaten their comparators in places, and the working assumption in this literature is that the potassium ion does the work whatever the anion is. The two studies with a modern randomised design that pair a potassium salt with hydroxyapatite both used citrate, and the larger was funded by a toothpaste manufacturer⁸⁹, so a claim about the nitrate pairing cannot borrow their results.

Does adding hydroxyapatite mean giving up fluoride?

Not in principle, though it usually does in practice. Four of the six UK products combining a potassium salt with hydroxyapatite carry no fluoride. The formulation obstacle is real, because calcium and fluoride can react in the tube, and the usual answer is to choose a fluoride salt that does not, which is why monofluorophosphate keeps appearing on combination labels. What a fluoride-free paste gives up is decay protection, not sensitivity relief.