Skip to content

Chapter 9: Sugar — How Often, Not Just How Much

In which we learn that the pattern of sugar consumption matters alongside the amount, that the two are far harder to separate than I used to claim, and that the one you can actually change is the one worth aiming at

The Tooth Fairy, exhausted by the endless parade of sweets.

"Cut back on sugar."

You've heard this advice your entire life. Dentists say it. Parents say it. Public health campaigns say it. And it's not wrong, exactly. Sugar does contribute to tooth decay. Reducing sugar intake is generally good for oral health.

But the conventional framing leaves out half the picture: how often you consume sugar matters alongside how much, and for most people it is the half they can actually do something about.

I need to open this chapter with a correction, because an earlier version of it overstated the case and some of you will have read that version. I used to tell you flatly that frequency matters more than quantity. The honest position is more interesting than that, and I'd rather hand you the real argument than a tidy one.

This chapter will still reframe your understanding of sugar's role in dental disease. By the end you'll think differently about that piece of cake — and you may enjoy it more, knowing that when you eat it genuinely changes what it costs you.

The Stephan Curve

In 1944, a dental researcher named Robert Stephan made a foundational discovery.1 He measured the pH in dental plaque after volunteers rinsed with glucose solutions, and he plotted the results over time.

What he found was consistent and reproducible:

  1. Immediate drop: Within minutes of sugar exposure, plaque pH fell sharply—often to 5.0 or below (well under the critical 5.5 threshold).

  2. Minimum reached: pH reached its lowest point approximately 5-20 minutes after exposure.

  3. Gradual recovery: Over the next 30-60 minutes, pH slowly rose back toward baseline as saliva buffered the acid and bacteria metabolized the remaining sugar.

This pattern became known as the Stephan Curve, and it remains foundational to understanding caries development.

Modern measurement has sharpened those numbers. In a 2024 double-blind crossover study, a sucrose challenge drove plaque pH to 5.1 at seven minutes, brought it back above the 5.5 threshold at around twelve minutes, and still had not returned to baseline thirty minutes later.15 That trio is worth holding on to: the damage window is shorter than people fear, and the full recovery is longer than they assume.

The key insight: each sugar exposure triggers an acid attack lasting 20-60 minutes. During this window, demineralization is occurring. Between attacks, when pH is above critical threshold, remineralization can happen.

Your dental fate depends on the balance between these phases.

The Mathematics of Frequency

Let's do some simple math.

Single daily exposure: One dessert after dinner. One acid attack, lasting approximately 30-40 minutes. That's about 40 minutes of demineralization per day, and 23+ hours of potential remineralization.

Three daily exposures: Breakfast includes fruit juice, lunch includes soda, dinner includes dessert. Three acid attacks, totaling perhaps 100-120 minutes of demineralization, with substantial recovery time between.

Continuous exposure: A sweetened coffee sipped throughout the morning, a few candies in the afternoon, a soda with dinner, a snack before bed. Five or six separate exposures, or—worse—continuous exposure during sipping. pH rarely rises above critical threshold. Demineralization dominates the day.

The total amount of sugar might be similar across these scenarios. The time spent below the dissolution threshold is not remotely similar, and that is the quantity that dissolves teeth.

That arithmetic is sound. Now let me complicate it honestly, because the epidemiology is messier than the arithmetic.

Science Note — what the evidence actually says about frequency versus amount

This is the most-contested claim in the book and I have been on the wrong side of the nuance. Here is the real state of it.

The evidence for amount is stronger than the evidence for frequency. The cleanest test is a model that adjusts each for the other. A longitudinal study following 1,702 Finnish adults across three surveys found that the amount of sugar intake — but not the frequency — was significantly associated with DMFT over follow-up.5 A NHANES analysis of 10,514 US adults found amount survived adjustment for frequency, while frequency measured as items per day did not survive adjustment for amount.6 The WHO-facing systematic reviews that set the "below 10% of energy" guidance are built on amount, and their 2022 update does not reframe around frequency at all.7

The evidence for frequency is weaker than its popularity. A 2026 systematic review of paediatric studies did conclude frequency is the stronger determinant — but 34 of its 37 studies were cross-sectional, the synthesis was qualitative with no meta-analysis, and it carries no GRADE certainty rating.8 That is vote-counting across heterogeneous studies, not a head-to-head test. I'm not going to cite it as vindication.

Why the question is partly ill-posed. In real human diets, amount and frequency are tightly correlated — people who eat sugar often also eat more of it. Untangling them statistically is genuinely hard, and anyone claiming a clean separation is overselling.9

What survives, and it's the useful part. The same review that makes the collinearity argument draws a conclusion I find far more practical than "which one wins": the relationship is not symmetrical. Cutting the amount while keeping the frequency high does not appear to work. Cutting the frequency tends to drag the amount down with it. And frequency is the more tangible target — "stop grazing" is a thing a person can actually do, where "reduce your intake to 8% of energy" is not.

And a mechanistic explanation for why the data are muddy. An agent-based simulation of plaque biofilm over 60 days found three regimes: at low total intake the biofilm stayed healthy regardless of frequency; at high total intake it went cariogenic regardless of frequency; and frequency was decisive only at intermediate totals — which is exactly where average human consumption sits.10 If frequency only matters in the middle of the range, population studies averaging across the whole range will show amount dominating. That is a model, not data, but it is the best available account of the discrepancy.

So: frequency is the lever, amount is the load. Both matter. Frequency is where most people have leverage, and reducing it reduces the other one too.

One more finding, and it deserves to sting a little: when fluoride is used properly, the relationship between sugar consumption and caries becomes very low or absent.9 The single most effective thing in this chapter is not a dietary change at all.

Historical Evidence: The Hopewood House Study

One of the most compelling demonstrations of this principle comes from the Hopewood House Study, conducted in Australia in the 1950s.

Hopewood House was a children's home where, due to philosophical beliefs of the founders, children were raised on a vegetarian diet extremely low in refined sugar.2 The diet wasn't designed for dental research—it was an ideological choice—but researchers took advantage of the natural experiment.

The children of Hopewood House had remarkably low rates of tooth decay compared to the general population. Their teeth weren't inherently different; their genetic backgrounds were typical for Australian children of the era. What was different was their sugar exposure pattern: nearly absent.

When some children left Hopewood House and adopted normal Australian diets, their cavity rates rose to match the general population within a few years.

The study suggested that sugar exposure below a certain threshold resulted in minimal caries, while exposure above that threshold caused disease. The relationship wasn't strictly linear—there seemed to be a threshold effect where the repair system could handle low levels of challenge but became overwhelmed above a certain frequency.

The Vipeholm Study: Ethics and Evidence

The most definitive—and ethically troubling—study on sugar consumption patterns was the Vipeholm Study, conducted in Sweden from 1945 to 1953.3

Researchers at the Vipeholm institution — a Swedish hospital for people with severe intellectual disabilities — fed different groups of residents controlled diets with varying sugar amounts and patterns over several years. Some received sugar with meals only. Others received sugar between meals. Others received sticky toffees that adhered to teeth for extended periods.

The findings were clear:

  1. Sugar consumed with meals caused minimal additional caries beyond baseline.
  2. Sugar consumed between meals substantially increased caries.
  3. Sticky sugars consumed between meals caused the most caries of any pattern tested.

It's not the sugar in your food. It's the sugar between your food.

Where this evidence came from

I am not going to pass over this quickly, because the data in the section above came at a cost somebody else paid.

The Vipeholm subjects were institutionalized people with severe intellectual disabilities. They did not consent and could not have. They were deliberately fed diets designed to rot their teeth, for years, and many of them lost teeth as the intended outcome of the experiment. The study was partly funded with industry interest in demonstrating that sugar could be consumed safely.

This would not be approved today, anywhere, by anyone. "The data were useful" is not a defence, and the honest thing is to say that the foundation of the frequency argument rests on an experiment that should never have happened. I cite it because the finding is real and because pretending we learned it some other way would be its own dishonesty. You should know whose teeth bought it.

Science Note — Vipeholm wasn't as tidy as it's usually reported

A 2021 re-reading of the Vipeholm data11 makes a point the textbook summaries drop: the variation between individuals was enormous. Some participants on the most aggressively cariogenic regimen developed no new lesions at all.

That isn't a footnote — it's the thing most sugar advice gets wrong. Sugar exposure is a risk factor operating on wildly different substrates. Salivary flow and composition, existing microbial community, fluoride exposure, enamel formed well or poorly in childhood: all of it modifies the dose-response. Vipeholm is often taught as a clean demonstration that pattern X produces outcome Y. It's better read as a demonstration that pattern X shifts the odds, and that individual susceptibility is large enough to swamp the effect in some people.

Which is why this book keeps refusing to give you a single number. The person who gets cavities on a careful diet is not imagining it, and the person who has never had one is not virtuous.

Why Frequency Matters: The Biological Logic

The Stephan Curve explains why frequency matters so much. Each sugar exposure triggers an independent acid attack. More frequent exposures mean more attacks, less recovery time, and net demineralization.

But there's more to the story.

Biofilm Metabolism

The bacteria in dental plaque don't metabolize sugar instantly. When you eat sugar, you're feeding a microbial community that takes time to absorb, process, and excrete acidic byproducts. A single large sugar dose saturates this metabolic capacity relatively quickly—the bacteria can only work so fast.

Frequent small doses, however, keep the bacteria in constant production mode. Each new sugar arrival restarts the metabolic clock. The total acid produced may actually be greater from many small exposures than from one large one.

Bacterial Selection

Remember the ecological plaque hypothesis? Frequent acid stress selects for acid-tolerant species like S. mutans. If you eat sugar once a day, the acid environment is transient—there's a lot of time at neutral pH, favoring a diverse community.

If you consume sugar continuously, the environment is persistently acidic. Acid-tolerant species gain competitive advantage. The community shifts toward organisms that produce more acid. A self-reinforcing cycle develops.

Frequency doesn't just cause more acid attacks—it reshapes the ecosystem toward one that causes even more acid attacks.

Salivary Exhaustion

Saliva provides buffering, remineralization minerals, and antimicrobial factors. But these resources aren't unlimited. Continuous acid challenges may partially exhaust buffering capacity. The minerals deposited during remineralization take time to accumulate in saliva.

Spacing out sugar exposures allows salivary resources to replenish between challenges.

The "Free Sugars" Distinction

Nutritional guidelines increasingly distinguish between free sugars and intrinsic sugars:

Free sugars are monosaccharides and disaccharides added to foods by manufacturers, cooks, or consumers, plus sugars naturally present in honey, syrups, and fruit juices.

Intrinsic sugars are sugars that are naturally present within the structure of whole fruits and vegetables.

This distinction matters for dental health because:

  1. Free sugars are more readily available to bacteria. They dissolve quickly in saliva and reach plaque rapidly.

  2. Intrinsic sugars are released more slowly as the food matrix is chewed and digested. The sugar release is spread over time rather than arriving as a bolus.

  3. Whole fruits contain fiber that stimulates saliva and provides mechanical cleaning. The chewing required extends exposure time but also activates defenses.

This doesn't mean whole fruit is decay-proof—someone who constantly snacks on grapes will still experience acid challenges. But whole fruit consumed in normal meal patterns is far less cariogenic than equivalent sugar from juice or candy.

Practical Applications

Good Patterns

Sugar with meals, and with no gap: Consuming sweets as part of a meal rather than separately. The other foods provide buffering, stimulate saliva, and dilute the sugar. This is the best-evidenced advice in the section and the detail matters: in human plaque-pH studies, eating non-sugary food before, during or after sugared coffee markedly reduced the pH fall — and a pause between the sugary and non-sugary parts abolished the benefit.13 The sweet eaten inside the meal is cheaper than the identical sweet eaten in the gap afterwards.

Consolidated treats: If you're going to have candy, eat a reasonable amount at one time rather than grazing. One acid attack from a candy bar should beat six from six individually unwrapped candies.

Tooth Fairy Note — on advice I cannot actually prove

I want to flag that last one, because I've repeated it confidently for years and I went looking for the study behind it.

There isn't one. Nobody has run the experiment that holds total sugar constant and compares one sitting against the same amount spread across a day. It is a reasonable extrapolation from the Stephan curve and I still believe it. But it is an extrapolation, and you are entitled to know which of my instructions are measured and which are merely sensible.

Two places it may actually mislead you:

  • If the sitting is enormous, the biofilm modelling suggests high total intake drives dysbiosis regardless of how you arrange it. "All at once" is not a licence.
  • For sour sweets the advice may invert. Erosion scales with contact duration, and a pile of sour candy eaten in one long sitting is one long acid bath. This advice is much better justified for chocolate than for anything labelled extreme.

What I can defend without hedging is the version with a meal attached: eat it in one sitting, with food, and don't leave a gap. Two of those three components have direct human measurement behind them.

Sweet beverages consumed quickly: If you drink soda (which I'd generally discourage), drink it over a short period rather than sipping throughout the day.

Bad Patterns

Continuous sipping of sweetened beverages: Coffee with sugar throughout the morning. Sports drinks during extended exercise. Soda nursed over an afternoon. Each sip restarts the acid clock.

Hard candies and lozenges: These dissolve slowly in the mouth, providing continuous sugar release. A single hard candy might deliver sugar for 15-20 minutes—an extended acid attack from a tiny amount of sugar.

Dried fruits as constant snacks: Raisins, dates and other dried fruits are sugar-concentrated, and snacking on them all day is a genuine frequency problem. But not for the reason you've been told — see the Science Note below, because the stickiness part turns out to be wrong.

Nighttime sugar: Any sugar before bed gets extended access to your teeth during the low-saliva overnight period.

Sugar in beverages retained in the mouth: Adding sugar to tea or coffee that you hold in your mouth while working extends contact dramatically.

Science Note — sticky is not the same as retained

Every dentist tells you to avoid sticky foods. Someone finally measured it, and the advice does not survive contact with the data.

Researchers took twenty-one commercial foods, asked consumers to rate how sticky each one felt, and then measured how much of each was actually retained on human teeth and how fast salivary sugar cleared.12 There was no correlation with perceived stickiness. On either measure.

What was most retentive? Cookies, crackers and potato chips. What was among the poorest retained? Caramels, jelly beans, raisins and milk chocolate bars. Chocolate-caramel bars showed high initial retention and then cleared very fast.

I find this genuinely funny. The toffee that feels welded to your molars is often gone before the cracker crumbs are, and the crisps nobody thinks of as a dental problem sit in your fissures being quietly fermented.

Two honest caveats: retention is a surrogate for cariogenicity, not a caries outcome, and this is essentially one careful study rather than a literature. But it is direct human measurement, and it should make you suspicious of any food ranking built on how things feel in the mouth.

The practical upshot isn't "eat caramels." It's that starchy savoury snacks deserve a place on the bad-patterns list that they are almost never given, and that "avoid sticky foods" is the wrong rule. The right rule is still about the clock, not the texture.

Chewing Gum: The Inversion

Here is the one case where frequency runs the other way — and the credit belongs somewhere slightly different from where it usually goes.

Chewing sugar-free gum after eating is well supported. A meta-analysis of twelve intervention studies found a preventive fraction of 28% against non-chewing controls.14 The mechanism is unglamorous: chewing makes saliva, saliva clears sugar and buffers acid. It is the only post-sugar countermeasure that attacks clearance rather than chemistry, which is precisely the variable this chapter is about.

Xylitol specifically is a sugar alcohol that tastes sweet but cannot be effectively fermented by S. mutans. The bacteria take it up, attempt to metabolize it, waste energy, and produce no acid. That mechanism is real and the surrogate evidence is decent: against sorbitol gum, xylitol reduced S. mutans counts in 12 of 14 head-to-head comparisons and plaque in 6 of 10 studies.

But on caries specifically, it managed 3 of 5 trials, and xylitol-only gum came out at 33% preventive fraction against plain sugar-free gum's 28% — overlapping intervals.14 The Cochrane review of ten RCTs and 5,903 participants found a single positive result, at high risk of bias.4

So chew gum after meals, and let it be xylitol gum if you like — it probably has a small edge and it certainly does no harm. But the chewing is the part you can count on, and the delivery form matters more than the molecule: gum studies are consistently positive where xylitol candy studies mostly are not.

Either way, the inversion holds: more frequent chewing is beneficial rather than harmful. You are repeatedly stimulating the one fluid that does the repairing.

The Permission Structure

I want to be clear about what this chapter is and isn't saying.

I'm not saying sugar is fine in unlimited quantities. Total sugar intake does matter, and the metabolic health consequences of excessive sugar go far beyond dental disease.

I'm not saying you should eat dessert with every meal. Concentrated sugar is still a challenge; it's just a more manageable challenge when consolidated.

I am saying that the pattern matters more than you've been told, and that it's the part you can change. If you're going to consume sugar — and most people will — understanding the frequency effect lets you reduce harm without necessarily reducing pleasure. And because frequency and amount travel together, cutting the grazing tends to cut the total without your having to count anything.

That birthday cake at a party? Eat it. Enjoy it. It's one acid attack, occurring with a meal, likely followed by hours without additional sugar.

That candy jar on your desk that you nibble from throughout the day? That's where the damage accumulates.

Garden Rule — count the occasions, not the grams

Nobody sustains a gram count. Almost anyone can count occasions.

  • Aim at how many times a day sugar enters your mouth, not at how much. It's the measurable one, and it pulls the total down behind it.
  • Attach sweets to meals, with no gap before or after.
  • Stop drinking sugar slowly. A soda over twenty minutes is one attack; the same soda sipped across an afternoon is a siege. Sipping is the single worst pattern in this chapter.
  • Chew sugar-free gum afterwards when you can.
  • Nothing sweet after the last brushing of the day. Overnight, flow falls by 80–90% and whatever you left behind has hours of undisturbed access.
  • Brush twice daily with fluoride toothpaste regardless, because it outperforms every dietary change in this chapter and it is the one thing here with Tier 1 evidence behind it.

The most important change most people can make isn't eliminating sugar. It's restructuring when they consume it — and being honest that the amount is coming along for the ride.

Now let's examine the structure that makes bacteria so difficult to dislodge: the biofilm, or plaque. Understanding this fortress is essential to understanding why mechanical disruption matters so much.

  1. Stephan, R. M. (1944). Intra-oral hydrogen-ion concentrations associated with dental caries activity. Journal of Dental Research, 23(4), 257-266. ↩

  2. Harris, R. (1963). Biology of the children of Hopewood House, Bowral, Australia. 4. Observations on dental-caries experience extending over five years (1957-61). Journal of Dental Research, 42(6), 1387-1399. ↩

  3. Gustafsson, B. E., et al. (1954). The Vipeholm dental caries study; the effect of different levels of carbohydrate intake on caries activity in 436 individuals observed for five years. Acta Odontologica Scandinavica, 11(3-4), 232-264. ↩

  4. Riley, P., Moore, D., Ahmed, F., Sharif, M. O., & Worthington, H. V. (2015). Xylitol-containing products for preventing dental caries in children and adults. Cochrane Database of Systematic Reviews, 3, CD010743. DOI: 10.1002/14651858.CD010743.pub2. Ten RCTs, 5,903 participants. The only positive finding was for fluoride toothpaste containing 10% xylitol versus fluoride alone (~13% caries reduction over 2.5–3 years), explicitly described as low-quality evidence to be interpreted with caution due to high risk of bias. For lozenges, tablets, syrups and wipes the confidence intervals were compatible with both reduction and increase. See also Söderling, E., & Pienihäkkinen, K. (2025). BMC Oral Health. DOI: 10.1186/s12903-025-06602-1 — 31 studies; xylitol gum beat sorbitol gum on mutans streptococci in 12 of 14 comparisons and on plaque in 6 of 10, but on caries in only 3 of 5 trials. ↩

  5. Bernabé, E., Vehkalahti, M. M., Sheiham, A., Lundqvist, A., & Suominen, A. L. (2016). The shape of the dose-response relationship between sugars and caries in adults. Journal of Dental Research, 95(2), 167-172. DOI: 10.1177/0022034515616572. Longitudinal, 1,702 dentate Finnish adults across three surveys. In a mutually adjusted linear mixed effects model, the amount of — but not the frequency of — sugars intake was significantly associated with DMFT throughout follow-up. The amount–caries association was weaker, though not eliminated, in adults using fluoride toothpaste daily. ↩

  6. Alosaimi, N., & Bernabé, E. (2022). Amount and frequency of added sugars intake and their associations with dental caries in United States adults. International Journal of Environmental Research and Public Health, 19(8), 4511. DOI: 10.3390/ijerph19084511. NHANES 2011–2016, n = 10,514. Amount survived adjustment for frequency (DMFS RR 1.11, 95% CI 1.07–1.15 per SD); frequency as items/day did not survive adjustment for amount (RR 1.02, 0.99–1.04). Cross-sectional design caps the strength of inference. ↩

  7. Moores, C. J., Kelly, S. A. M., & Moynihan, P. J. (2022). Systematic review of the effect on caries of sugars intake: ten-year update. Journal of Dental Research, 101(9), 1034-1045. DOI: 10.1177/00220345221082918. And the original: Moynihan, P. J., & Kelly, S. A. M. (2014). Journal of Dental Research, 93(1), 8-18. DOI: 10.1177/0022034513508954. These form the WHO evidence base and are built on amount rather than frequency. ↩

  8. Pombo-Lopes, J., Silva, M. L., Neves, B., & Grillo-Evangelista, J. (2026). Impact of sugar consumption on dental caries in children and adolescents: a systematic review of frequency versus quantity. BMC Public Health. DOI: 10.1186/s12889-026-27780-9. Concludes frequency is the stronger determinant, but 34 of 37 included studies were cross-sectional, synthesis was qualitative with no meta-analysis, and no GRADE certainty rating was applied. ↩

  9. van Loveren, C. (2019). Sugar restriction for caries prevention: amount and frequency. Which is more important? Caries Research, 53(2), 168-175. DOI: 10.1159/000489571. Narrative review. Source of both the collinearity argument and the reciprocity argument — that reducing amount without reducing frequency does not appear effective while the reverse does — and of the observation that where fluoride is appropriately used, the relation between sugar consumption and caries is very low or absent. ↩↩

  10. Head, D., Devine, D. A., & Marsh, P. D. (2017). In silico modelling to differentiate the contribution of sugar frequency versus total amount in driving biofilm dysbiosis in dental caries. Scientific Reports, 7, 17413. DOI: 10.1038/s41598-017-17660-z. Agent-based plaque biofilm model, 60 simulated days. At low total intake (≤12.5 g/L/d) the biofilm stayed non-cariogenic regardless of frequency; at high total intake (≥22.5 g/L/d) it was cariogenic regardless of frequency; frequency was decisive only at intermediate totals. The authors describe their predictions as physiologically plausible rather than definitive. ↩

  11. Vieira, A. R. (2021). Individual susceptibility to dental caries: the Vipeholm study. Monographs in Oral Science. DOI: 10.1159/000520766. A re-reading emphasizing the enormous between-participant variability, including participants on the most cariogenic regimens who developed no new lesions. ↩

  12. Kashket, S., Van Houte, J., Lopez, L. R., & Stocks, S. (1991). Lack of correlation between food retention on the human dentition and consumer perception of food stickiness. Journal of Dental Research, 70(10), 1314-1319. DOI: 10.1177/00220345910700100101. Twenty-one commercial foods; consumer stickiness ratings showed no correlation with measured retention on human teeth or with salivary sugar clearance. Cookies, crackers and potato chips were most retentive; caramels, jelly beans, raisins and milk chocolate bars among the poorest retained. ↩

  13. Rugg-Gunn, A. J., Edgar, W. M., & Jenkins, G. N. (1981). The effect of altering the position of a sugary food in a meal upon plaque pH in human subjects. Journal of Dental Research, 60(5), 867-872. DOI: 10.1177/00220345810600050201. The pH fall from sugared coffee was considerably reduced by eating non-sugary foods before, during or after — and this favourable action was negated by a pause between the sugary and non-sugary courses. ↩

  14. Newton, J. T., Awojobi, O., Nasseripour, M., et al. (2020). A systematic review and meta-analysis of the role of sugar-free chewing gum in dental caries. JDR Clinical & Translational Research, 5(3), 214-223. DOI: 10.1177/2380084419887178. Twelve intervention studies. Sugar-free gum versus non-chewing controls: preventive fraction 28% (95% CI 7–48%). Xylitol-only gum: 33% (4–61%). The authors note considerable heterogeneity and moderate trial quality. ↩↩

  15. Mankar, S., Amaechi, B. T., Kanthaiah, K., et al. (2024). Clinical, Cosmetic and Investigational Dentistry, 16, 287-297. DOI: 10.2147/CCIDE.S472427. Double-blind crossover, 16 subjects. A sucrose challenge drove plaque pH to 5.1 ± 0.1 at seven minutes, back above the 5.5 threshold at approximately twelve minutes, and not to baseline for more than thirty minutes. ↩