Chapter 40: Candy Season, Cold Season

In which the calendar turns against your teeth twice in quick succession, and I make the case that the second half is worse than the first — because nobody warns you about it, and because the instructions on the packet actively tell you to do the wrong thing

---

There is a stretch of the year, roughly October through February in the northern half of the world, when my workload reliably increases.

It begins with a single evening of organized sugar distribution to children. Everyone knows about that one. Dentists write newspaper columns about it. Parents negotiate treaties over it. It is, as it happens, the part of the season I worry about least.

What follows is the part nobody writes columns about: four months of people sucking slowly on sweetened, acidified tablets, eight or ten a day, for weeks at a time, on the explicit instruction of the packaging — while congested, breathing through their mouths, and taking medicines that turn off their saliva.

Let me take the famous one first, and then the one that actually fills my boxes.

Halloween

I want to tell you something about Halloween that I find genuinely remarkable.

Nobody has ever studied it.

I don't mean the evidence is thin or conflicting. I mean that the single most concentrated sugar event in the American calendar — an evening that every dentist in the country has an opinion about, that generates an annual tide of advice columns and dental-office poster campaigns — has never been the subject of a published study on dental outcomes. No caries incidence data. No consumption volumes in the peer-reviewed literature. No evaluation of a candy buy-back program, though thousands of dental practices run them. Nothing.

There are exactly two papers in the medical literature with Halloween in the title that touch this territory. One is about how well handwashing removes SARS-CoV-2 from candy wrappers. The other is a gut-microbiome commentary that uses trick-or-treating as a metaphor.

That absence is more interesting than a weak finding would have been, and it should make you appropriately suspicious of anyone — me included — who speaks confidently about what Halloween does to teeth. What we have instead is the general sugar literature from Chapter 9, applied by inference. Which is legitimate. It is just not the same thing as knowing.

So here is what the inference actually supports.

The pillowcase is not the problem. The desk drawer is.

One evening of enthusiastic candy consumption is, in Stephan-curve terms, one long acid exposure. It is not nothing. But it resolves.

What does the damage is what happens to the rest of the haul. A pillowcase of sweets gets rationed — two or three pieces a day, for five or six weeks, usually between meals, often at a desk or in front of a screen. That is thirty-odd separate acid exposures spread across a month and a half, arriving precisely in the gaps where nothing else is buffering.

The evening is a single event. The aftermath is a habit. The aftermath is what I would change.

Garden Rule — let them eat it, then let it end

My honest advice to a parent, knowing that it is inference and not measurement:

• Don't ration it across weeks. Rationing feels responsible and is dentally the worst thing you can do with a large quantity of sweets. It converts one exposure into thirty.
• Attach it to meals. Dessert after dinner, not a mid-afternoon treat — and with no gap between the meal and the sweet, because in human plaque-pH studies a pause abolishes the protective effect of the surrounding food.¹
• Let the rest go. Buy-back, freeze it, donate it, bin it. No study supports this; it follows directly from the point above.
• Sort out the sour ones. Of everything in that bag, the deliberately sour sweets are the ones with a genuinely different mechanism — see below. If you are going to confiscate a category, confiscate that one.
• Brush that night with fluoride toothpaste and don't make the evening a reason to skip it.

And keep some proportion. One Halloween does not decide anything. A pillowcase rationed into February, every year, is a pattern — and patterns are what I collect teeth from.

The sour ones are a different problem

Most sweets feed bacteria, which then make acid. Sour sweets cut out the middleman and deliver the acid directly.

The pH values that circulate in dental education — Sour Skittles at 2.2, Pixy Stix 1.9, Fun Dip 1.8, one sour spray product measured at 1.6 — come from a practitioner review in a state dental association journal rather than a peer-reviewed measurement paper.² I would treat the specific decimals as approximate. The magnitude is not in doubt: at pH 1.6 to 2.2 these products are as acidic as the contents of your stomach — and unlike stomach acid, which reaches your teeth only diluted and rarely, they are held deliberately in the mouth for the sensation.

Two findings sharpen this.

Science Note — why pH alone underestimates a sour sweet

Titratable acidity predicts erosion better than pH does. When researchers exposed enamel to slurries of original and sour versions of the same candies, lesion depth correlated with titratable acidity — how much total acid there is to neutralize — and only weakly with the starting pH.³ A pH reading tells you the acid's strength at one instant. Titratable acidity tells you how long it can keep attacking as saliva tries to buffer it, which is the question that matters.

And saliva may not rescue you. The same work found that while saliva plausibly protects against the original-flavour versions, it is "much less likely to protect against the erosive effects of sour candies."

A correction worth making, because you will encounter the claim: citric acid is not especially erosive because it chelates calcium. That mechanism was tested directly, across 27 experimental groups, and the authors concluded chelation has no relevant influence on enamel dissolution.⁴ Citric acid is more erosive than its pH alone suggests — but the reason is its large reservoir of undissociated acid, which keeps re-supplying protons at the tooth surface long after a simple strong acid at the same pH would have been neutralized.

The practical consequence of all this: for sour sweets, duration is the dose. Enamel hardness falls measurably after two minutes of exposure and keeps falling out to two hours.⁵ This is the one category where my usual "eat it in one sitting" advice works against you — a pile of sour candy consumed in one long sitting is one long acid bath.

And then the sweets come with a Drug Facts panel

Here is where I get genuinely exercised, and where the evidence turns out to be much better than I expected.

From roughly November onward, a large fraction of the adult population begins consuming hard candy on a medical schedule. They do not think of it as candy. It is in the pharmacy aisle, it has an active ingredient, and the box gives dosing instructions. It is candy.

The labelling gap is not an accident of design

In the United States, cough drops and throat lozenges are regulated as drugs. They therefore carry a Drug Facts panel rather than a Nutrition Facts panel — which means the sugar content is not disclosed at all. The most you are likely to see is a voluntary "10 calories per drop."

In the UK and EU, medicines regulation runs the other way: excipient-warning rules force the manufacturer to state the exact quantity of sugar per lozenge. Which is why every hard number I am about to give you is British.

Read this next table as the numbers that exist for products on both sides of the Atlantic, disclosed on one side only.

Product (UK Summary of Product Characteristics)	Sucrose	Glucose	Total sugars	Added acid
Strepsils Original	1.496 g	1.013 g	2.51 g	tartaric
Strepsils Honey & Lemon	1.44 g	0.98 g	2.42 g	tartaric
Strepsils Orange + Vitamin C 100 mg	1.44 g	0.97 g	2.41 g	tartaric + ascorbic
Strepsils Strawberry Sugar Free	—	—	0	tartaric

The Original's label says it plainly: "This medicine contains 1.496 g sucrose per lozenge and 1.013 g glucose per lozenge."⁶

For the food-labelled drops, sugar is declared by weight rather than per unit: Fisherman's Friend menthol-eucalyptus at 88 g sugars per 100 g — the most sugar-dense product I found, in one of the smallest lozenges, working out near a gram per drop. Halls menthol in the EU at 81 g per 100 g, which on a 2.8 g drop is about 2.3 g of sugar each. Ricola Original at 76 g per 100 g, and Ricola publishes no drop weight, so no honest per-drop figure exists.⁷

Three-quarters to seven-eighths sugar by weight. In a dosage form engineered, by design, to dissolve slowly against your teeth.

The instructions mandate the worst possible pattern

This is the part that moved me from irritation to something closer to indignation.

Look at what the packets actually tell you to do. Halls Mentho-Lyptus: "Repeat every 2 hours as needed" — no daily maximum stated. Chloraseptic: "1 lozenge every 2 hours as needed" — no daily maximum. Cepacol: "may be repeated every 2 hours as needed."⁸

Every two hours. All day. And not one of those labels mentions teeth.

Chapter 9 spends considerable effort persuading you that the number of daily sugar occasions is the variable you can control. Here is an entire product category whose printed instructions specify an occasion roughly every two waking hours, in a form designed for prolonged contact, during an illness that may itself have dried your mouth out. If you set out to design a cariogenic regimen, you would struggle to improve on it.

The zinc lozenge deserves its own warning

The sharpest example I found, and it is the product people choose specifically because it is the evidence-based cold remedy.

One popular zinc acetate lozenge carries, on its own Supplement Facts panel: 20 calories, 5 g total carbohydrate, 4 g sugar, with dextrose as the first other ingredient. Its directions: "Completely dissolve in mouth one (1) lozenge every 2 wakeful hours, not to exceed 8 lozenges per day."⁹

Do the arithmetic the label invites. Eight lozenges. Four grams of dextrose each. Thirty-two grams of sugar, delivered in eight separate slow-dissolving installments across a single day — on the authority of a printed instruction that never once mentions your teeth.

To be fair to zinc: it does have real evidence behind it for shortening a cold. But it also has a well-documented adverse-event profile — pooled across 28 trials and 5,446 participants, non-serious adverse events ran 41% higher than placebo, a number-needed-to-harm of 7, dominated by unpleasant taste and mouth irritation.¹⁰ One trial found 63% of zinc users reporting adverse effects against 31% on placebo.

If you want zinc, ask for a formulation without the dextrose. The zinc is the medicine. The four grams of sugar are the delivery vehicle, and there is no reason you should have to accept them.

"Sugar-free" is the hook, not the reassurance

You might reasonably conclude from all this that the answer is to buy the sugar-free version. It is better. It is not the solution people believe, and the evidence here is the best in the chapter.

Start with the US definition. "Sugar free" on an American label means under 0.5 grams of sugars per serving — a threshold that is nearly meaningless applied to a two-gram item.

Then the mechanism. Sugar alcohols are usually described as non-cariogenic, but the foundational statement of that position is more careful than its popular version: sorbitol, mannitol, maltitol and lactitol are all slowly fermented to acid by oral bacteria, and acid formation from them "can be too slow to overcome the buffering power of plaque and saliva."¹¹ Only xylitol is essentially non-fermentable.

Read that conclusion again and notice the assumption buried in it. Too slow to overcome buffering — which holds only if the exposures are intermittent, with buffering windows between them. Hold a slowly-fermented sugar alcohol in your mouth every two hours all day and you have removed the very condition the reassurance depends on.

And isomalt — slowly fermentable, not xylitol — is the sweetener in essentially every sugar-free medicated lozenge on the market.

Science Note — the case that should change your mind about sugar-free lozenges

A patient with full-mouth restorations, normal salivary flow, and no other obvious risk factor developed extensive ditching around the margins of brand-new restorations within a single year. The only unusual thing in the history: 8 to 12 "sugar-free" cough drops per day.

The investigators then took the drops into the lab. Streptococcus sobrinus fermented their isomalt and sorbitol to a terminal pH of 4.2 — against 3.7 for glucose, so not as bad as sugar, but comfortably below the 5.5 dissolution threshold. It reached pH 5.1 at modest concentrations, and consumed about 71% of the available isomalt over 62 hours. The authors' conclusion: patients should be cautioned against chronic overuse of "sugar-free" cough drops.¹²

This is a case report plus single-species in vitro work — Tier 3, and I will not pretend otherwise. But it is corroborated from an adjacent direction: a 2024 case describes root caries on virtually every posterior tooth, despite excellent oral hygiene, after twenty years of sugar-free nicotine lozenges sweetened with mannitol.¹³

The lesson is not that sugar alcohols are secretly sugar. It is that "slowly fermented" is a promise that expires with prolonged residence time, and a lozenge is a dosage form whose entire purpose is prolonged residence time.

There is also the companion trap: removing the sugar does not remove the acid. When 97 paediatric formulations were measured, the sugar-free ones were not less erosive than the sugar-containing ones, and 57% of all of them had a pH below 5.5. The better predictor of erosive potential turned out to be the dosage form, with effervescent tablets the worst.¹⁴ Strepsils Sugar Free still carries tartaric acid, exactly like the sugared versions.

A genuine counterexample, because blanket warnings are lazy

I am not going to tell you that every medicated lozenge is a sugar bomb, because one of them is quietly well designed.

Cepacol Extra Strength is sugar-free on isomalt, maltitol and sucralose, carries no acidulant at all, and contains sodium bicarbonate — which is pH-raising. It is, chemically, closer to the rinse in Chapter 12 than to a sweet.

Compare Luden's Wild Cherry, the worst-designed product I came across: sucrose and corn syrup, carrying three separate acids — ascorbic, citric and malic. A full cariogenic load and a triple erosive challenge in one slowly dissolving tablet.

Same aisle. Same shelf. Same two-hour dosing instruction. Wildly different consequences for your teeth, and nothing on either box tells you which is which.

One more reason to read the ingredients rather than the brand: Cold-Eeze is not one formulation. Three current labels share the brand and the active ingredient. Two contain corn syrup, honey and sucrose. The third, marked sugar-free, is on acesulfame-K, glycine and isomalt. "I take Cold-Eeze" tells you nothing whatsoever about sugar exposure.

Getting Through a Cold Without Paying in Enamel

Read the ingredients, not the front	You are looking for two things: a fermentable sugar (sucrose, glucose, dextrose, corn syrup, honey) and an acid (citric, malic, ascorbic, tartaric). The best products have neither; the worst have both.
Prefer a sugar-free lozenge with no acidulant	Rarer than it should be, and worth hunting for. A bicarbonate-containing one is better still.
If it must be sugared, don't graze it	The label says every two hours. Your teeth would prefer fewer, and nothing terrible happens if you take it less often than the maximum.
Warm salt water is free and does the same job	For a sore throat, the evidence for a lozenge's analgesic benefit over a warm saline gargle is not strong enough to justify eight sugar exposures a day. A gargle costs nothing and has no pH problem.
Ask for zinc without the dextrose	Tablets and capsules exist. The zinc is the medicine.
Keep brushing, twice, with fluoride toothpaste	Non-negotiable, and more protective than every other row in this table combined. Being ill is the worst possible time to let this slip, and it is exactly when people do.
Rinse with water after any lozenge	Free, harmless, and it shortens the clearance tail.

The dry mouth nobody warned you about

Now the other half of the season, which is the half I expected to be easy to write and turned out to be the most humbling research of the lot. I went looking for evidence that a winter cold dries your mouth out and damages your teeth. Most of what I had believed turned out to be unsupported, and two pieces of it were flatly wrong.

Let me take the four mechanisms in order, from the one with real evidence to the one I have had to throw away.

The medicine, and which medicine

This part holds up, but only for a specific class.

The physiology is the key and it is routinely got wrong. Salivary fluid secretion is driven by the parasympathetic system, through acetylcholine at muscarinic receptors — chiefly M3.²⁰ Sympathetic stimulation drives protein release, not volume. Which means:

• First-generation antihistamines — diphenhydramine, chlorpheniramine — are potent muscarinic antagonists. They have a direct mechanism for reducing saliva volume, and chlorpheniramine is the one where somebody actually measured it: significantly reduced salivary flow against placebo in a double-blind crossover.²¹ Diphenhydramine, remarkably, has never had its effect on salivary flow measured at all — the expectation is pharmacological, not observed.
• Pseudoephedrine is a sympathomimetic. It has no receptor-level mechanism for reducing saliva volume. The systematic review of medications causing salivary dysfunction rates its evidence as weak, resting on a single narrative review of a combination product from which pseudoephedrine's contribution can't be separated.²² It may thicken saliva and reduce glandular blood flow. "Pseudoephedrine shuts off your saliva" is not what the receptor physiology says.
• Phenylephrine does not appear in that systematic review at all. Nor does oxymetazoline. Entirely unstudied. (Phenylephrine has separately been proposed for removal from US over-the-counter products on the grounds that it doesn't work as a decongestant, which rather overtakes the question.)

Science Note — the honest chain, and where it breaks

Link one is solid. Sustained low salivary flow demineralizes teeth. In an in situ study, 85% of subjects with unstimulated flow at or below 0.16 mL/min exceeded mean mineral loss, against 33% at 0.17–0.30 and none at all above 0.30 mL/min.²³ That is a clean dose-response.

Link two is real but small. Anticholinergic load measurably reduces flow: each 1 pmol/mL rise in serum anticholinergic activity corresponded to a 0.058 mL/min fall in unstimulated flow.²⁴

Link three does not exist. Nobody has studied over-the-counter cold-medicine use and caries, or cold medicine and enamel. Not once. And there is a scale mismatch in the chain: every demineralization study concerns chronic hyposalivation over weeks to years, while the only antihistamine effect anyone measured was a single dose peaking at three hours. A week of cold medicine is not a year of polypharmacy.

So the defensible sentence is this: first-generation antihistamines measurably reduce salivary flow, and sustained low flow is a demonstrated cause of demineralization — but nobody has measured what a week of cold medicine does to enamel. If you are on them for a season of hay fever rather than a week of flu, the chain is more worrying. If you are on them for four days, I would not lose sleep.

One further gap worth knowing: the large meta-analysis of medications and dry mouth produced pooled odds ratios for urological drugs (5.91), antidepressants (4.74) and psycholeptics (2.59) — and no pooled estimate for antihistamines or decongestants at all, because the studies aren't there.²⁵

And the thing I did not expect: the cold medicine isn't only drying your mouth. It is a sugar delivery vehicle too. In a double-blind randomized crossover, all nine paediatric oral liquid medications tested produced a drop in biofilm pH statistically indistinguishable from 10% sucrose.²⁶ A separate analysis of 30 products found 70% had a pH below 5.5, and sucrose content ranging up to 75.6% — with 43% of products above the 10% threshold generally considered cariogenic.

The mouth breathing: real, smaller than the lore, and not from allergies

Mouth breathing is genuinely associated with more plaque and more gingival inflammation. A 2026 systematic review of eleven studies found increased gingival bleeding, plaque accumulation and gingival alterations — while being careful to call its own evidence "limited," declining to perform a meta-analysis, and finding caries associations only for initial lesions and the most advanced ones.²⁷ The best single study adjusted for gender, crowding and plaque quantity, and found mouth breathing's effect survived — localized to the palatal surfaces.²⁸

The strongest evidence that the dryness is doing the damage, rather than just the plaque, is the one randomized trial in this entire area: a salivary substitute applied to the anterior sextants as an adjunct to scaling and root planing outperformed scaling alone on gingival index and bleeding.²⁹ That runs the causal arrow the right way.

But keep the magnitude honest. In a study of 257 preschoolers, mouth breathing carried an adjusted prevalence ratio of 1.57 for anterior caries — and in the same model, sucrose frequency carried 4.02.³⁰ Diet outweighed breathing mode more than twofold.

Science Note — two things I had to delete from this chapter

Allergic rhinitis does not cause cavities. I was going to write that it did. A 2026 systematic review pooling 60 studies and 343,975 participants across 22 countries, with two-sample Mendelian randomisation alongside it, found no association between allergic rhinitis and dental caries: OR 1.01 (95% CI 0.86–1.18).³¹ In the same paper, asthma was associated (OR 1.44, with genetic support) and atopic dermatitis weakly so (1.09) — so the analysis was capable of detecting an effect. It simply wasn't there for rhinitis.

This is corroborated by the only direct test of the mechanism: in 40 children with allergic rhinitis against 40 healthy controls, there was no significant difference in caries experience, unstimulated salivary flow, or salivary buffering capacity.³² You will find large database studies claiming the opposite — they measure dentist visits, not examined teeth, in children who see doctors more often for everything.

And the "dehydrated plaque is more virulent" mechanism is not a finding. You will meet it in textbooks. It originates in a single discussion sentence in a 1990 paper on open bite in children — a study in which plaque index did not differ significantly between groups, and which measured no hydration, no plaque pH, no water content and no airflow. The authors offered it as a speculation. Nothing in thirty-five years has tested it. It became fact by repetition, which is exactly the process this book exists to interrupt.

What also doesn't exist: any measurement of how fast a mouth dries during sustained mouth breathing. The only number in the literature is a calculation — evaporative loss can reach roughly 0.21 mL/min during mouth breathing, "normally much less" — and the same physiologist notes that the anterior hard palate carries the thinnest salivary film in the mouth, thin enough that evaporation can take it to zero.³³ Which is a satisfying fit with the palatal localisation above. It is also two unrelated studies joined by my own inference, and you should treat it as such.

One further wrinkle, from the only study to put a hygrometer near the teeth of mouth-breathing volunteers: it found no significant difference in intraoral humidity between anterior and posterior teeth. Exhaled air is warm and saturated, so mouth breathing alternates dry inhaled air with wet exhaled air. The drying is real and localized and probably modest — and its true magnitude is unknown.

Do not tape your mouth shut

Since the internet will suggest it: mouth taping has been reviewed, across ten studies and 213 patients total. Two showed any improvement in apnoea measures and the rest showed nothing, most of the included studies specifically excluded people with nasal obstruction, and the reviewers flagged a risk of serious harm including asphyxiation in exactly the congested population this chapter is about.³⁴

There is also no dental or salivary endpoint anywhere in that literature. Nobody has measured what taping does to a mouth, only to breathing.

If you are congested enough that you can't breathe through your nose at night, the answer is treating the congestion, not sealing the alternative.

The furnace, not the humidifier

Here is the one where I had it backwards, and I have corrected Chapter 2 accordingly.

I assumed dry winter air dries your mouth. The reasoning is so intuitive that I never questioned it. Then I went looking.

Humidifying indoor air has no demonstrated effect on dryness symptoms. A Cochrane review of 13 studies and at least 4,551 participants found little to no effect on dryness of the eyes, skin or upper respiratory tract — and dry mouth was not even among the outcomes measured. Perceived stuffiness actually increased with humidification.³⁵ Bedroom humidification for dry mouth has never been trialled at all.

What has been measured is the heating. In a study of 50 volunteers measured four times each across winter and summer, with separate parotid and submandibular collection: parotid and submandibular flow were significantly higher in winter than summer, room heating significantly lowered parotid flow, and air conditioning almost completely abolished the seasonal effect.³⁶ Cold air raises salivary flow; a controlled crossover at 4°C, 21°C and 37°C found resting flow significantly increased in the cold.³⁷

And the one study that measured room humidity against sialometry directly found the association running the opposite way from the folklore: higher humidity, less unstimulated saliva, with winter showing the highest flow at the lowest humidity.³⁸ It is a small study and I would not build a doctrine on it. But it is the only one that looked, and it does not say what I assumed.

No caries study of any design has ever used ambient humidity or heating season as an exposure. That line is simply unstudied, and I have struck it from this book rather than keep repeating a plausible thing.

Garden Rule — turn the heat down, skip the appliance

• If you wake with a dry mouth in winter, look at the heating before the humidity. Turn the thermostat down overnight, or redirect the vent away from the bed. Warmth is the measured variable.
• A humidifier is fine if it makes your nose and throat feel better. It has no demonstrated benefit for your mouth, and it may make the room feel stuffier.
• Water by the bed, and sip it if you wake. Dull, free, and it addresses the actual deficit.
• Don't tape your mouth.

And the season itself?

Since the whole chapter rests on the idea that this time of year is worse for teeth, I should tell you what the seasonal epidemiology actually shows. It is thinner than you'd imagine.

Caries incidence seasonality has never been studied with modern methods. The entire literature is three papers, the most recent from 1985, two of them without abstracts.

Dental emergency data is real but contradicts itself. A time-series of 3.88 million outpatient visits in Shanghai found risk rising below about 7°C, with cumulative risk ratios at −3°C of 2.40 for pulpitis and 1.92 for oral disease overall, strongest in people over 60.³⁹ A European series found the opposite sign — visits rising above 21°C and falling during cold spells. And the best-designed study of the three, covering a full year of odontogenic abscesses, found no significant correlation with temperature, humidity, precipitation, sunshine, season or lunar phase. Do not let anyone — including me — tell you this is settled.

Periodontal measurements do vary by season, and the finding is a lesson in reading effect sizes. Across 5,908 patients and 128,309 teeth, probing depths were significantly higher in summer than winter. The authors then stated plainly that the differences were under 0.5 mm and of no clinical relevance, and suggested that year-round food availability, climate control and vitamin D supplementation have flattened whatever seasonality once existed.⁴⁰

One condition genuinely does cluster, and nobody knows why. Acute necrotising ulcerative gingivitis: in 684 cases from a six-year series, 55% presented in summer and 8% in winter. A second series on another continent found 72% clustering in two distinct windows. Both single-centre, neither adjusted for school terms or nutrition. It is a real pattern with no explanation attached.

And the folk etymology that turns out to be earned: the cold sore. A three-year prospective study of 149 student nurses found upper respiratory infection rate among the strongest predictors of recurrent herpes labialis, and the canonical laboratory model reactivates latent HSV-1 with transient hyperthermia alone, with infectious virus detectable within 14 hours.⁴¹ Fever blister is an accurate name. What has never been shown is a calendar pattern — the link is to being ill, not to the month.

The children's-medicine problem

Everything above applies with more force to children, who get their medicines as syrups.

The foundational observation is old and blunt: children on long-term sucrose-sweetened medicines had a mean decayed/filled score of 5.55 against 1.26 in controls.¹⁵ The methodology is dated, but the direction has held up across four decades of controlled observational work. In chronically ill children taking liquid medicines, caries odds run roughly three times those of comparison groups — 77.8% with decay against 52.7%.¹⁶

The confounding here is severe and I want to be honest about it: chronic illness brings its own dietary changes, its own hygiene difficulties, and frequently its own drug-induced dry mouth. No randomized trial is possible or ethical. So the defensible statement is that it is now widely accepted that sugar-containing medicines contribute to caries in chronically sick children — not that medicines cause caries.

What genuinely surprised me is the structural gap. Forty-five years of literature on this question, and no systematic review or meta-analysis has ever been conducted. The most recent synthesis is explicitly narrative.

If your child is on a long-term liquid medicine

This is the one section of this chapter I would actually raise with a professional rather than handle at home.

• Ask whether a sugar-free formulation exists. Often one does, and nobody mentions it because nobody asks.
• Ask whether it can be given at a time that allows brushing afterwards — and specifically not as the last thing before sleep, which is the traditional slot and the worst one.
• Tell the dentist what the child is taking. Long-term medication changes the risk calculation and may justify more frequent reviews, fluoride varnish, or sealants.
• Never stop or re-time a prescribed medicine on my account. I am a fairy with opinions about enamel. Their prescriber is managing the illness.

One genuinely startling example of how invisible this is: a widely sold UK children's cough syrup lists sucrose as an active ingredient — 1.7 g per 5 mL dose, up to four doses daily, which is 13.6 g of sucrose a day at the label's own maximum. Its regulatory document carries a warning for diabetics and no warning about teeth at all.¹⁷

The vitamin C footnote

Since this is the season for it: chewable and effervescent vitamin C is acidic, and it is acidic by design.

Chewable vitamin C tablets have been measured at pH 2.3, and a 500 mg chewable dose can drop salivary pH below the 5.5 threshold and hold it there for up to 25 minutes.¹⁸ Across six children's chewable supplements, every one was acidic (pH 2.99–4.77), enamel weight loss tracked inversely with pH, and pure vitamin C products eroded more than multivitamins — while products containing minerals eroded less, which is the calcium-saturation principle showing up again.

The defensible claim is that chewable vitamin C can erode enamel and has done so in documented cases — not that vitamin C supplements cause erosion. If you want the vitamin, a swallowed tablet bypasses the problem entirely. If you want the chewable, don't hold it in your mouth, and rinse afterwards.

What we genuinely do not know

I have tried throughout this book to mark the edges of the evidence, and this chapter has unusually many. Seven things in it are simply unstudied:

• Nobody has surveyed how many lozenges people actually take during an illness. The "eight a day" figure above is arithmetic from a label, not observed behaviour.
• No one has run plaque pH telemetry on a lozenge or a hard candy dissolving in the mouth. The chain from prolonged clearance to extended acid exposure to demineralization is mechanistically sound and consistent across every piece of evidence we have — but it has never been assembled in a single study on cough drops.
• No cohort study, case-control study or trial has ever tested cough-drop use as a caries exposure. Three case reports are the entire body of direct human evidence.
• Acesulfame-K appears in a great many sugar-free lozenges and is absent from the oral-health literature altogether. Not fermented by oral bacteria as far as anyone has checked — which is less reassuring than it sounds, because nobody has checked.
• Zinc's oral adverse effects are pooled into a general "non-serious" category. Nothing separates mouth effects from stomach effects, and there is nothing at all on zinc lozenges and the oral microbiome.
• Menthol has no published minimum inhibitory concentration against any oral bacterium, and no trial of menthol alone on plaque or gingivitis. It is in oral products as a flavour.
• Halloween, as discussed, has never been studied at all.

Tooth Fairy Note — on the cruelty of the menthol

One last thing, because it is the sort of detail that makes me want to sit down.

Menthol does stimulate saliva — genuinely, through the cold receptor on your tongue, which is a small mechanical kindness in the middle of a cold.¹⁹ But a study of a cold-water-and-menthol protocol for dry mouth found that it was not effective at relieving the dryness, and was effective at reducing the feeling of thirst.

Menthol makes a dry mouth feel better without making it any wetter.

Which is, if you want a single image for this entire chapter, the business model: a sweetened, acidified tablet that soothes the sensation of a sore throat while quietly dissolving the teeth behind it, on a dosing schedule printed by someone who was thinking about your throat and had no reason to think about anything else.

I am not angry at the lozenge. It was designed to comfort you and it does. I am angry that in four months of every year, millions of people follow a printed instruction eight times a day without anyone ever having mentioned what it costs — and then arrive in a dentist's chair in March, having done everything they were told.

---

Next, the practical synthesis: how to assemble everything in this book into something you will actually do.

---

---

1. 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 favourable effect of surrounding food was negated by a pause between the sugary and non-sugary courses. ↩

2. Loewen, R., Marolt, J., & Ruby, D. (2008). Pucker up: the effects of sour candy on your patients' oral health. A review of the dental erosion literature and pH values for popular candies. Northwest Dentistry, 87(2), 20-21, 24-25, 28-29. PMID: 18491469. Note that this is a practitioner review in a state dental association journal, not a peer-reviewed measurement study, and carries no DOI. Its values are the most widely cited in the field and are broadly consistent with other sources, but the specific decimals should be treated as approximate. ↩

3. Wagoner, S. N., Marshall, T. A., Qian, F., & Wefel, J. S. (2009). In vitro enamel erosion associated with commercially available original-flavor and sour versions of candies. Journal of the American Dental Association, 140(7), 906-913. DOI: 10.14219/jada.archive.2009.0284. Lesion depth correlated positively with titratable acidity and only negatively with initial pH. In vitro, with a 25-hour exposure that is not physiological. ↩

4. Azadi-Schossig, P., Becker, K., & Attin, T. (2016). Chelating effect of citric acid is negligible for development of enamel erosions. Clinical Oral Investigations, 20(7), 1577-1587. DOI: 10.1007/s00784-015-1634-x. Superfusion model, 27 experimental groups. Citric acid was more erosive than hydrochloric acid at matched pH, but the authors conclude that chelating effects do not have a relevant influence on hydroxyapatite dissolution; the driver is the interplay of pH, titratable acidity and buffering capacity. See also Hughes, J. A., et al. (2000). Journal of Dentistry, 28(2), 147-152. ↩

5. Naidoo, L. S., Onwubu, S. C., Murugan, N., & Singh, S. (2021). Examining the effect of exposure time on the erosive potential of sour candy. South African Dental Journal, 76(2). DOI: 10.17159/2519-0105/2021/v76no2a2. Fifty human tooth specimens; Vickers hardness fell significantly at every timepoint from 2 minutes to 2 hours, with Raman spectroscopy confirming demineralization through loss of the 960 cm⁻¹ phosphate peak. ↩

6. Summaries of Product Characteristics, UK Electronic Medicines Compendium (medicines.org.uk/emc). These are the manufacturers' own regulated disclosures, which is the strongest provenance available for a product figure, but they are labels rather than studies. Note that the SmPCs do not state total lozenge weight, so percentage by weight is not calculable. ↩

7. On-pack nutrition panels as transcribed in Open Food Facts. Crowd-sourced from photographs of labels; weaker provenance than a manufacturer's regulated document. Per-drop figures depend on lozenge weight, which varies between listings for the same product — Fisherman's Friend appears as both 1.15 g and 1.25 g, shifting the per-drop sugar between roughly 1.01 g and 1.11 g. Where no drop weight is published, as with Ricola, no per-drop figure should be calculated. ↩

8. Directions panels as published on DailyMed (dailymed.nlm.nih.gov). US cough drops and throat lozenges are regulated as drugs and therefore carry a Drug Facts panel, which does not require disclosure of sugar content. "Sugar free" on a US label means less than 0.5 g sugars per serving (21 CFR 101.60©(1)). ↩

9. NIH Dietary Supplement Label Database, record 232155. Supplement Facts and Directions as printed by the manufacturer. ↩

10. Hunter, J., Arentz, S., Goldenberg, J., et al. (2021). Zinc for the prevention or treatment of acute viral respiratory tract infections in adults: a rapid systematic review and meta-analysis of randomised controlled trials. BMJ Open, 11(11), e047474. DOI: 10.1136/bmjopen-2020-047474. Twenty-eight RCTs, 5,446 participants; non-serious adverse events RR 1.41 (95% CI 1.17–1.69), number-needed-to-harm 7. See also Nault, D., Machingo, T. A., Shipper, A. G., et al. (2024). Zinc for prevention and treatment of the common cold. Cochrane Database of Systematic Reviews, 5, CD014914. DOI: 10.1002/14651858.CD014914.pub2 — 34 RCTs, 8,526 participants, non-serious adverse events RR 1.34 (1.15–1.55), moderate-certainty evidence. And Hemilä, H., et al. (2020). BMJ Open, 10(1), e031662 — 63% of zinc users reported adverse effects against 31% of placebo users. None of these reviews reports oral adverse events separately from gastrointestinal ones. ↩

11. Edgar, W. M., & Dodds, M. W. (1985). The effect of sweeteners on acid production in plaque. International Dental Journal, 35(1), 18-22. PMID: 3858227. Sorbitol, mannitol, maltitol and lactitol are all slowly fermented to acid by oral bacteria; xylitol is virtually non-fermentable. The authors' observation that acid formation from sugar alcohols "can be too slow to overcome the buffering power of plaque and saliva" assumes intermittent exposure. ↩

12. Mayo, J. A., & Ritchie, J. R. (2009). Acidogenic potential of "sugar-free" cough drops. Open Dentistry Journal, 3, 26-30. DOI: 10.2174/1874210600903010026. Case report plus in vitro fermentation. S. sobrinus 6715 fermented the drops' isomalt and sorbitol to terminal pH 4.2 (glucose gave 3.7), reached pH 5.1 at 50 mg extract/mL, and consumed approximately 71% of the isomalt over 62 hours. ↩

13. Ehrhardt, C., Mäkinen, K. K., & Cobb, C. M. (2024). Case Reports in Dentistry, 2024, 6635130. DOI: 10.1155/2024/6635130. Multiple root caries lesions despite excellent oral hygiene after more than twenty years of sugar-free nicotine lozenges sweetened with mannitol (75.7%) plus maltodextrin. At the far extreme of the same continuum, see Mandel, L., & Carunchio, M. J. (2011). Journal of the American Dental Association, 142(4), 406-409 — complete destruction of the dental crowns of a 19-year-old after three years of sucrose-containing fentanyl lozenge misuse. ↩

14. Maguire, A., Baqir, W., & Nunn, J. H. (2007). Are sugars-free medicines more erosive than sugars-containing medicines? An in vitro study of paediatric medicines with prolonged oral clearance used regularly and long-term by children. International Journal of Paediatric Dentistry, 17(4), 231-238. DOI: 10.1111/j.1365-263X.2007.00826.x. Ninety-seven formulations; sugar-free mean pH 5.73 against sugar-containing 5.26 (p > 0.05, i.e. no significant difference in erosive potential); 55 of 97 had pH below 5.5; dosage form was a more significant predictor of erosive potential than sugar content, with effervescent tablets the worst. ↩

15. Roberts, I. F., & Roberts, G. J. (1979). Relation between medicines sweetened with sucrose and dental disease. British Medical Journal, 2(6181), 14-16. DOI: 10.1136/bmj.2.6181.14. Mean decayed/filled score 5.55 in children on long-term sucrose-sweetened medicines against 1.26 in controls. Methodologically dated. ↩

16. Fatih, M. T., et al. (2025). Pediatric liquid medications and dental caries: a narrative review. Health Science Reports, 8(7), e71115. DOI: 10.1002/hsr2.71115. Covers 33 studies; reports OR 3.142 (p < 0.001) for caries in 455 chronically ill children on paediatric liquid medicines (77.8% with decay against 52.7%). Note that this is explicitly a narrative review — no systematic review or meta-analysis of this question exists. The association is confounded by the underlying illness, diet, hygiene difficulties and drug-induced xerostomia, and no randomized trial is possible. ↩

17. Summary of Product Characteristics, UK Electronic Medicines Compendium, product 2062. Sucrose is listed among the active ingredients at 1.7 g per 5 mL dose; dosing is 5 mL (ages 1–5) or 10 mL (5 and over), three to four times daily. The document carries a warning for patients with diabetes and no dental warning. Its stablemate for chesty coughs (emc 1479) contains no sucrose. ↩

18. Bahal, P., & Djemal, S. (2014). Dental erosion from an unexpected source. Case Reports in Dentistry, 2014, 485387. DOI: 10.1155/2014/485387. Chewable vitamin C tablets measured at pH 2.3; a 500 mg dose dropped salivary pH below 5.5 for up to 25 minutes. For the product range across children's chewables, see Lertsooksawat, W., et al. (2024) — six products, all acidic at pH 2.99–4.77, enamel weight loss inversely correlated with pH (r = −0.806, p = 0.029). All in vitro or case-report level. ↩

19. Houghton, J. W., Carpenter, G., Hans, J., Pesaro, M., Lynham, S., & Proctor, G. (2020). Agonists of orally expressed TRP channels stimulate salivary secretion and modify the salivary proteome. Molecular & Cellular Proteomics, 19(10), 1664-1676. DOI: 10.1074/mcp.RA120.002174. Mouth rinsing with menthol, a TRPM8 agonist, increased whole-mouth saliva flow and total protein secretion compared with unstimulated saliva. Small human experimental study with surrogate outcomes. On menthol's lack of demonstrated antimicrobial activity in the mouth, see Potocka, W., Assy, Z., Bikker, F. J., & Laine, M. L. (2023). Molecules, 28(20), 7178. DOI: 10.3390/molecules28207178, which reports no minimum inhibitory concentrations and no biofilm-inhibition data for menthol, noting that it appears in oral products chiefly as a flavour and fragrance. ↩

20. Proctor, G. B., & Carpenter, G. H. (2007). Regulation of salivary gland function by autonomic nerves. Autonomic Neuroscience, 133(1), 3-18. DOI: 10.1016/j.autneu.2006.10.006. Salivary fluid secretion is parasympathetic, mediated by acetylcholine at muscarinic receptors (chiefly M3); sympathetic noradrenergic stimulation drives protein release rather than volume. This distinction is why antimuscarinic agents have a direct mechanism for reducing saliva volume and sympathomimetics do not. ↩

21. Vargas, R., McMahon, F. G., & Ryan, J. R. (1989). Effects of terfenadine, chlorpheniramine and atropine on salivary flow. Journal of International Medical Research, 17(2), 141-146. DOI: 10.1177/030006058901700207. Double-blind crossover, 21 adults, single dose, sialometry at baseline and 3 hours. Chlorpheniramine significantly reduced mean salivary flow versus placebo (p < 0.01); atropine more strongly (p < 0.001); terfenadine was indistinguishable from placebo. The published abstract does not give mL/min values. For the magnitude of a reference antimuscarinic, see Brion, N., Beaumont, D., & Advenier, C. (1988). British Journal of Clinical Pharmacology, 25(1), 27-32 — atropine 1 mg reduced salivary output by 2.25 ± 0.36 mL from a control value of 4.17 ± 0.42 mL (p < 0.001). ↩

22. Wolff, A., Joshi, R. K., Ekström, J., et al. (2017). A guide to medications inducing salivary gland dysfunction, xerostomia, and subjective sialorrhea: a systematic review sponsored by the World Workshop on Oral Medicine VI. Drugs in R&D, 17(1), 1-28. DOI: 10.1007/s40268-016-0153-9. PRISMA; 3,867 records screened, 269 analysed, graded by evidence level. Diphenhydramine, chlorpheniramine and pseudoephedrine all sit in the weak-evidence category; phenylephrine and oxymetazoline do not appear at all. The authors' own stated limitation is worth quoting: "While xerostomia was a commonly reported outcome, objectively measured salivary flow rate was rarely reported." ↩

23. Bardow, A., Nyvad, B., & Nauntofte, B. (2001). Relationships between medication intake, complaints of dry mouth, salivary flow rate and composition, and the rate of tooth demineralization in situ. Archives of Oral Biology, 46(5), 413-423. DOI: 10.1016/S0003-9969(01)00003-6. Twenty-eight subjects, root-surface demineralization by quantitative microradiography over 62 days in situ. 85% of subjects with unstimulated flow ≤ 0.16 mL/min exceeded mean mineral loss, against 33% at 0.17–0.30 mL/min and none above 0.30 mL/min. ↩

24. Michail, A., Almirza, M., Alwaely, F., & Arany, S. (2023). Anticholinergic medication burden and salivary flow in older adults. Archives of Oral Biology, 156, 105824. DOI: 10.1016/j.archoralbio.2023.105824. Cross-sectional, 55 adults. ↩

25. Tan, E. C. K., Lexomboon, D., Sandborgh-Englund, G., Haasum, Y., & Johnell, K. (2018). Medications that cause dry mouth as an adverse effect in older people: a systematic review and metaanalysis. Journal of the American Geriatrics Society, 66(1), 76-84. DOI: 10.1111/jgs.15151. Fifty-two studies included, 26 pooled. Pooled odds ratios: urological 5.91, antidepressants 4.74, psycholeptics 2.59. No pooled estimate was possible for antihistamines or decongestants. ↩

26. Alencar, K. P., Peixoto, D. F., Máximo, F. D. N., Farias, I. A. P., & Sampaio, F. C. (2024). Cariogenic potential of pediatric oral liquid medications: a randomized crossover in situ study. Brazilian Oral Research, 38, e107. DOI: 10.1590/1807-3107bor-2024.vol38.0107. Double-blind randomized crossover, 12 subjects, plaque pH by microelectrode over 30 minutes; all nine medications produced pH drops with no statistically significant difference from 10% sucrose. For product composition, see Akolgo, G. A., et al. (2026). BMC Oral Health, 26, 1426. DOI: 10.1186/s12903-026-09011-0 — 30 products, pH 2.76–7.57 with 70% below 5.5, sucrose content 0–75.6%. ↩

27. Kimura, A. C. R. S., de Arruda, J. A. A., Drumond, V. Z., Martins-Júnior, P. A., Mesquita, R. A., & Abreu, L. G. (2026). Oral health status of mouth-breathing children and adolescents: a systematic review. International Journal of Paediatric Dentistry, 36(1), 24-32. DOI: 10.1111/ipd.70022. Eleven included studies, all cross-sectional; no meta-analysis was performed, so no pooled odds ratio exists. The authors characterize their own evidence as limited. ↩

28. Wagaiyu, E. G., & Ashley, F. P. (1991). Mouthbreathing, lip seal and upper lip coverage and their relationship with gingival inflammation in 11-14 year-old schoolchildren. Journal of Clinical Periodontology, 18(9), 698-702. DOI: 10.1111/j.1600-051x.1991.tb00112.x. Cross-sectional, 201 children, two examiners. On multivariate analysis mouth breathing and lip coverage remained significant after adjustment for gender, crowding and amount of plaque, with mouth breathing's effect restricted to palatal sites. The only real attempt at adjustment in this literature, and it is thirty-five years old. ↩

29. Bhatia, A., Sharma, R. K., Tewari, S., & Narula, S. C. (2015). Effect of a salivary substitute on gingival health in mouth breathers. Journal of Oral Science, 57(3), 241-247. DOI: 10.2334/josnusd.57.241. Randomized, investigator-blind, n = 40. The only randomized trial in this area, and the best evidence that the dryness rather than the plaque is doing the damage. ↩

30. Soares, M. E. D. C., Ramos-Jorge, J., Lima, L. J. S., Moreira, L. V., Fernandes, I. B., Ramos-Jorge, M. L., & Galo, R. (2024). Mouth breathing and dental caries in preschool children. Brazilian Oral Research, 38, e057. DOI: 10.1590/1807-3107bor-2024.vol38.0057. Cross-sectional, 257 children aged 3-5. Anterior caries adjusted prevalence ratio 1.57 (1.01–2.46, p = 0.047); sucrose frequency in the same model 4.02 (2.03–7.95). Note the lower confidence bound clears 1 by 0.01. ↩

31. Yu, Q., Lan, L., Li, Y., He, Y., Hu, J., Fang, K., Ji, Y., Chen, Z., Gao, Y. D., & Zheng, Y. (2026). Association between asthma, allergic rhinitis, atopic dermatitis, and dental caries: evidence from systematic review with meta-analysis and Mendelian randomisation investigation. Journal of Global Health, 16, 04223. DOI: 10.7189/jogh.16.04223. PROSPERO CRD420251103903. Sixty studies, 343,975 participants, 22 countries; random-effects pooling plus two-sample Mendelian randomisation. Allergic rhinitis and caries: OR 1.01 (95% CI 0.86–1.18). Asthma: OR 1.44 (1.17–1.76) with Mendelian randomisation support. Atopic dermatitis: OR 1.09 (1.01–1.17). ↩

32. Wongkamhaeng, K., Poachanukoon, O., & Koontongkaew, S. (2014). Dental caries, cariogenic microorganisms and salivary properties of allergic rhinitis children. International Journal of Pediatric Otorhinolaryngology, 78(6), 860-865. DOI: 10.1016/j.ijporl.2014.03.001. Forty children with allergic rhinitis against 40 healthy controls, aged 6-13; no significant difference in DMFT/dmft, unstimulated salivary flow rate, buffering capacity, lactobacilli or sugar consumption. Small, and a negative finding rather than a refutation — but it is the only direct test of whether nasal obstruction lowers salivary flow in congested children. ↩

33. Dawes, C. (2004). How much saliva is enough for avoidance of xerostomia? Caries Research, 38(3), 236-240. DOI: 10.1159/000077760. A calculation and physiological synthesis rather than a measurement study: evaporative fluid loss from the mouth can only occur during mouth breathing, at a maximum of roughly 0.21 mL/min at rest and "normally much less." On the salivary film, see Won, S., Kho, H., Kim, Y., Chung, S., & Lee, S. (2001). Archives of Oral Biology, 46(7), 619-624, and its replication in Assy, Z., Jager, D. H. J., Brand, H. S., & Bikker, F. J. (2023). Clinical Oral Investigations, 27(2), 859-869 — the anterior palate carries the thinnest film and the lowest MUC5B. For the null finding on anterior versus posterior intraoral humidity in mouth-breathing volunteers, see Washino, T., Ikeda, M., Burrow, M. F., & Nikaido, T. (2025). Journal of Adhesive Dentistry, 27, 261-266. DOI: 10.3290/j.jad.c_2398. ↩

34. Rhee, J., Iansavitchene, A., Mannala, S., Graham, M. E., & Rotenberg, B. (2025). Mouth taping for sleep-disordered breathing: a systematic review. PLoS One, 20(5), e0323643. DOI: 10.1371/journal.pone.0323643. Ten studies, 213 patients total; two showed improvement in apnoea or desaturation measures, the remainder nothing. The authors flag a potentially serious risk of harm including asphyxiation in the presence of nasal obstruction, and note that most included studies excluded anyone with nasal obstruction. No dental or salivary outcome appears anywhere in this literature. ↩

35. Byber, K., Radtke, T., Norbäck, D., et al. (2021). Humidification of indoor air for preventing or reducing dryness symptoms or upper respiratory infections in educational settings and at the workplace. Cochrane Database of Systematic Reviews, 12, CD012219. DOI: 10.1002/14651858.CD012219.pub2. Thirteen studies, at least 4,551 participants. Primary outcomes were dry eye, dry nose, dry skin and upper respiratory infection — dry mouth was not an outcome. Dry nose OR 0.87 (0.53–1.42) and dry skin OR 0.66 (0.33–1.32), both null; perceived stuffiness increased with humidification. GRADE certainty low to very low throughout. The only study of humidification and dry mouth specifically is Hay, K. D., & Morton, R. P. (2006). Head & Neck, 28(9), 792-796 — a pilot in nine Sjögren's patients with no control group and no sialometry, delivering heated saturated air by nasal prongs rather than humidifying a room. ↩

36. Elishoov, H., Wolff, A., Kravel, L. S., Shiperman, A., & Gorsky, M. (2008). Association between season and temperature and unstimulated parotid and submandibular/sublingual secretion rates. Archives of Oral Biology, 53(1), 75-78. DOI: 10.1016/j.archoralbio.2007.08.002. Fifty healthy volunteers, each measured four times — twice in winter and twice in summer — with separate parotid and submandibular/sublingual collection. Winter secretion rates were significantly higher than summer (p < 0.02 parotid, p < 0.05 SM/SL); room heating in winter significantly lowered mean parotid flow; air conditioning almost completely cancelled the seasonal effects. The authors recommend that room temperature be recorded whenever salivary flow is assessed. ↩

37. Ligtenberg, A. J. M., Meuffels, M., & Veerman, E. C. I. (2020). Effects of environmental temperature on saliva flow rate and composition. Archives of Oral Biology, 109, 104593. DOI: 10.1016/j.archoralbio.2019.104593. Crossover, 20 subjects, randomized exposure to 4°C, 21°C and 37°C; cold significantly increased resting flow rate (p < 0.01) and protein and amylase output (p < 0.001), with no change in MUC5B concentration or viscosity. ↩

38. Ito, K., Nohno, K., Funayama, S., et al. (2026). Impact of temperature, humidity, dehydration, and psychological stress on salivary flow and xerostomia in young men: an observational study. PLoS One, 21(5), e0349221. DOI: 10.1371/journal.pone.0349221. Longitudinal, 39 healthy young men, four seasonal sessions, hygrometer at the point of collection. Unstimulated flow fell with temperature (β = −0.006 mL/min per °C, p = 0.008) and with humidity (β = −0.003 mL/min per 1% RH, p = 0.039). Winter showed the highest unstimulated flow (0.39 ± 0.22 mL/min) at the lowest humidity (46.5%). Xerostomia was associated with neither temperature nor humidity, only with a stress marker. Small, male-only, single-site, with temperature and humidity collinear across seasons — cite only as the one study that looked finding the association running the other way. ↩

39. Qian, Y., Gao, Y., Cai, B., Zhang, W., Wang, X., & Chen, R. (2022). Low ambient temperature as a novel risk factor of oral diseases: a time-series study. Science of the Total Environment, 812, 152229. DOI: 10.1016/j.scitotenv.2021.152229. 3,882,636 outpatient visits across all Shanghai public hospitals, 2016-2019. Cumulative risk ratios at −3°C: total oral diseases 1.92, pulpitis 2.40, periodontitis 1.62. Relative humidity was not included as an exposure or covariate. For the contradicting result, see Rieder, M., et al. (2026). Applied Sciences, 16, 6505. DOI: 10.3390/app16136505 — visits increasing above 21°C and cold spells reducing risk. For the null, see Ristow, O., et al. (2015). BMC Oral Health, 15, 21. DOI: 10.1186/s12903-015-0001-2 — 1,211 odontogenic abscess patients over a full year, no significant correlation with temperature, humidity, precipitation, sunshine, season or lunar phase. ↩

40. Peikert, S. A., Metzger, S., Kruse, A. B., et al. (2024). The seasonal occurrence of periodontitis — a retrospective cohort study from a practice-based research network. Clinical Oral Investigations, 28, 596. DOI: 10.1007/s00784-024-05972-0. 5,908 patients, 128,309 teeth. Probing depths significantly higher in summer than winter, with winter significantly lower than every other season — and the authors state the differences are under 0.5 mm and "of no clinical relevance," proposing that year-round food availability, climate control and vitamin D supplementation have flattened whatever seasonality once existed in industrialized countries. On the one condition that does cluster, see Arendorf, T. M., Bredekamp, B., Cloete, C. A., & Joshipura, K. (2001). Oral Diseases, 7(3), 150-154 — 684 cases of acute necrotising ulcerative gingivitis from a series of 19,944 periodontal patients, 55.4% presenting in summer against 8.4% in winter, with no established explanation. ↩

41. Friedman, E., Katcher, A. H., & Brightman, V. J. (1977). Incidence of recurrent herpes labialis and upper respiratory infection: a prospective study of the influence of biologic, social and psychologic predictors. Oral Surgery, Oral Medicine, Oral Pathology, 43(6), 873-878. DOI: 10.1016/0030-4220(77)90079-2. Three-year prospective study of 149 student nurses; 40-50% of the variance in recurrent herpes labialis incidence was explained by prior experience, upper respiratory infection rate, socioeconomic status and mood trait, with episode timing unrelated to menstrual cycle phase. For the laboratory model, see Sawtell, N. M., & Thompson, R. L. (1992). Journal of Virology, 66(4), 2150-2156 — transient hyperthermia efficiently reactivated latent HSV-1, with infectious virus detectable by 14 hours. No study establishes a calendar-seasonal pattern in recurrent herpes labialis itself. ↩