18  Bristles, heads, and the two-minute rule

Three small pieces of advice about the object itself. Two of them are Conditional, one is not really a recommendation at all, and the last is the most famous instruction in dentistry.

The advice

Toothbrush type: use a small toothbrush head, medium texture

Strength of recommendation: Conditional

— Delivering Better Oral Health, chapter 2, table 2a (1)

And, from the caries tables, the first bullet of the Strong brushing bundle for children aged 3 to 6:

  • on all tooth surfaces

— Delivering Better Oral Health, chapter 2, table 1b (1)

Notice what is not there. None of the graded recommendations in chapter 2 specifies how long to brush for. That gap is the most interesting thing in this chapter, and I have now got it wrong twice in different directions, so it is worth stating precisely.

The two-minute rule is not a graded recommendation in the summary tables. It is in the guideline: chapter 8 tells dental teams that thorough cleaning “may take at least 2 minutes,” refers elsewhere to “the recommended 2 minutes,” and discusses timers. What chapter 8 does not do is attach a strength label or a certainty rating to it, and what it gives as the reason is not a caries finding but a rationale about coverage: “The main rationale for this time period is to ensure that sufficient time is taken for all tooth surfaces to be cleaned effectively.”

Bristle stiffness

The citation is Ranzan and colleagues (2), a systematic review asking whether bristle stiffness and end-shape cause adverse effects on soft tissue. Note the framing: this is a review about harm, not benefit. Nobody is claiming a medium brush cleans better. The claim is that a hard one damages you.

Its conclusion:

Soft and extra-soft toothbrushes tend to be safer. Oral soft tissue injuries are similar for both tapered and end-rounded bristles. Further studies investigating adverse effects as a primary outcome are recommended.

— Ranzan and colleagues, 2019 (2)

The underlying finding, from one included study, is that hard brushes produced more gingival lesions than soft or medium at four and eight weeks, with no significant difference between medium and soft. And a limitation the review states plainly: only four of the included studies had adverse effects as their primary outcome. Everything else was measuring something else and noticed harm in passing.

So the evidence supports “not hard”. It does not distinguish medium from soft, and the review’s own conclusion leans toward soft while DBOH recommends medium. That is a small, real divergence, on low certainty evidence, in a Conditional recommendation. It is about as low-stakes as a disagreement can be, and I note it only because the book’s method requires noticing when the recommendation and its citation do not quite line up.

Head size

I could find no evidence statement in chapter 13 for the small-head element, and no citation. It travels with the bristle recommendation on the same line, at Conditional. It is a reasonable piece of clinical logic, a smaller head reaches further back and around corners, and it is untested.

The two-minute rule

Everybody knows you should brush for two minutes. Timers are built into electric toothbrushes, apps count it down, and the number is on the packaging.

It is not among the graded recommendations in Delivering Better Oral Health’s summary tables, though chapter 8 of the same guideline does advise it as a thoroughness heuristic, as set out above.

SIGN 138 explains why nobody grades it, in §5.5.2, and the passage is worth quoting in full because it is a model of how to write about an absence:

Toothbrushing for at least two minutes is recommended by the Childsmile programme, the SDCEP guidance and some toothpaste manufacturers. However, no studies were identified relating to duration of brushing and caries prevention.

There is insufficient evidence on which to recommend a specific duration for an episode of toothbrushing for the prevention of caries.

— SIGN 138, §5.5.2 (3)

That is exactly right, and it is exactly what a guideline should say when the evidence is not there. SIGN names who recommends it, states the gap, and declines to make a recommendation.

Where I overstated this

In my original blog post I wrote:

We all know about the (at least) two-minute rule, and it’s been encouraged by the Childsmile programme, the SDCEP guidance, dental professionals, and toothpaste manufacturers. But in fact, there is no evidence to support this.

The last sentence is too broad, and it is the error this book exists to object to.

There is no evidence on caries, which is what SIGN searched for and what the sentence should have said. There is evidence on plaque removal: longer brushing removes more plaque, which is unsurprising and has been shown in brushing-exercise studies of the sort discussed in Chapter 16.

Saying “no evidence” when you mean “no evidence for this outcome” is the same move as treating absence of evidence as evidence of absence, and I made it in a post complaining about people who make it. The distinction is not pedantic: a reader of my sentence would reasonably conclude that brushing duration does not matter, which is not what anyone has shown.

Why has nobody run the trial? Because it is unusually hard.

To test duration against caries you would need to randomize people to brush for fixed different lengths of time, verify adherence over years, and hold everything else constant, including how much toothpaste they use and how well they cover their teeth. Brushing time and brushing thoroughness are entangled: someone who brushes for two minutes almost certainly covers more surfaces than someone who brushes for thirty seconds, so a duration trial is really a thoroughness trial with a clock attached.

That is why the graded recommendation says “on all tooth surfaces” and the two-minute figure appears in chapter 8 as a rationale for coverage rather than as a recommendation in its own right. The guideline grades the thing that plausibly matters and offers the clock as a way of getting there. That is a good drafting decision and I want to credit it.

The two-minute rule is best understood as a usable heuristic for thoroughness, not a finding. Two minutes is roughly how long covering every surface takes. That is a perfectly respectable reason to keep saying it, provided nobody claims it was measured.

PubMed, 21 August 2026, for studies of toothbrushing duration with a caries outcome. None identified, consistent with SIGN’s finding. Searches for duration and plaque return results; I have not appraised that literature in detail, and this chapter’s claim about it is deliberately limited to its existence.

I have the full text of Ranzan 2019, supplied during writing.

Full record in appraisals/searches/ch18-brush-design.md.

Verdict

Certainty of evidence

Low for medium-texture bristles: a small review of harm, in which only four studies measured adverse effects as a primary outcome, showing hard is worse and medium indistinguishable from soft. This matches both Ranzan’s own GRADE table, which rates the bristle-stiffness evidence low after downgrading for risk of bias and indirectness while noting a dose-response gradient, and DBOH’s “low certainty evidence of gingival lesions when hard bristle brushes were used.” An earlier draft rated it very low, one level below both, without naming a domain that would justify the extra step. I have moved it back.

None for head size. None on caries for brushing duration, and no graded duration recommendation in the summary tables.

Directness to the advice as worded
Partial. Ranzan supports “not hard” rather than “medium”, and its own conclusion favors soft. Head size is uncited. The duration evidence that exists is about plaque, not caries.
Is the strength label defensible?

Yes. Conditional is right for low certainty evidence about a low-stakes choice, and declining to grade a brushing duration is exactly what the evidence supports.

This chapter’s correction is aimed at me, not at the guideline. On the two most famous pieces of toothbrush advice, DBOH is more careful than the profession that repeats them and more careful than I was: it declines to grade a duration, and where it does mention two minutes, in chapter 8, it gives coverage rather than caries as the reason.

What would change my mind
A trial with adverse effects as the primary outcome comparing soft against medium, which would settle whether DBOH or its own citation has the better of it. On duration, honestly, nothing: the trial is not worth running, and “cover every surface” is better advice than any number of seconds.

What this means for you

Do not use a hard toothbrush. That is the one part of this with evidence behind it, and hard brushes damaged gums in the studies that looked.

Beyond that, soft or medium, small head or slightly larger: nobody has shown it matters, and the guideline labels it Conditional accordingly.

Brush for about two minutes if it helps you. But the reason is not that two minutes was tested and found to be the right number; it was not. The reason is that covering every surface of every tooth takes roughly that long, and covering every surface is the actual instruction.