Letting a scalding cup of tea or coffee cool is a sensible precaution. A 2026 study of nearly one million UK adults found that people who said they preferred drinks very hot had about three times the relative risk of one kind of esophageal cancer compared with people who preferred them warm. The study did not measure what temperature any participant actually swallowed, and it cannot prove that cooling a cup prevents cancer. An earlier cohort that measured tea temperature directly adds weight to the concern about repeated heat exposure.
The esophagus is the tube that carries a drink from the throat to the stomach. The newer result concerned squamous cell carcinoma, which begins in the cells lining the tube. Researchers found no similar temperature pattern for the other major type, adenocarcinoma. That difference is important if you have seen a headline that speaks of esophageal cancer as one undivided disease.
You do not need to give up a favorite tea or coffee because of this research. The practical question is whether a drink is so hot that you hurry through tiny cautious sips or burn your mouth. Giving it time to become comfortable costs little and avoids a repeated heat exposure that may matter over years.
What the UK researchers followed
The 2026 UK hot-drinks study (https://pubmed.ncbi.nlm.nih.gov/42710538/) combined two existing health cohorts. The Million Women Study contributed 510,112 women, and UK Biobank contributed 467,170 adults. The total was 977,282 people. Participants reported whether they liked hot drinks warm, hot, or very hot, and how many cups of tea and coffee they usually drank. Researchers then used cancer records to identify new diagnoses over an average 14.2 years in the women's cohort and 11.1 years in UK Biobank.
Across both groups, 2,348 people developed esophageal cancer during follow-up. Of these, 1,045 had squamous cell carcinoma and 1,303 had adenocarcinoma. The distinction was available because cancer records included the tumor type. Without it, an association confined to one type might be diluted or mistakenly described as applying to both.
The researchers compared the self-reported temperature groups while accounting for drink frequency and other measured factors. For squamous cell cancer, the adjusted relative risk was 1.69 for people who preferred drinks hot versus warm, and 3.17 for those who preferred them very hot versus warm. The 95% confidence interval around the very-hot estimate was 2.49 to 4.03. A confidence interval describes the range of values compatible with the data under the model. It is not a prediction that any one person's risk will triple.
For adenocarcinoma, the corresponding very-hot estimate was 0.92, with a confidence interval from 0.76 to 1.12. That result did not show an increase. The two cancers have different patterns of risk, so a person with reflux should not infer that cooling coffee addresses the risks associated with reflux-related disease. The National Cancer Institute's esophageal cancer risk information explains the distinction.
The study also examined how often people drank tea or coffee. After accounting for preferred temperature, six or more cups a day versus fewer than six was associated with 1.71 times the relative risk of squamous cell cancer. This does not establish that six cups is a biological threshold. It was the dividing line chosen for analysis, and cup sizes and actual drinking temperatures were not verified at each serving. A daily habit may mean more repeated contact with hot liquid, but other habits can travel with a high cup count.
Three times the risk is not three times the number of cases
The relative figure sounds large because it compares one group with another. The cancer remained uncommon in these cohorts: 1,045 squamous cell cases occurred among 977,282 participants over roughly a decade or more. That rough fraction is about one case per thousand participants across the entire mixed population. It is not the exact personal risk for a warm-drink or very-hot-drink consumer, because their ages, follow-up times, and other risks differed.
The paper estimated that 14% of UK squamous cell cases might be avoided if people who preferred drinks very hot instead drank them hot. That was a model calculation assuming the association is causal. It was not an observed reduction after people changed their habits. The study did not test an intervention in which one group cooled its drinks and the other continued its usual routine.
Prospective follow-up is stronger than asking people with cancer to remember how hot they drank years earlier. It still leaves room for factors researchers could not measure perfectly. Smoking and alcohol are established risks for squamous cell cancer, and the authors adjusted for reported habits. The association remained across several subgroups and after excluding early diagnoses. These checks make simple explanations less likely, but they cannot turn a cohort study into a randomized trial.
What the study could not tell about temperature
Every UK participant chose a word for preferred temperature. No thermometer was placed in a cup at the moment of drinking. One person's hot may be another person's very hot, and adding milk or waiting between pours and sips changes the temperature in the mouth. The authors could not translate the UK labels into a particular number of degrees Celsius or Fahrenheit.
The International Agency for Research on Cancer classified drinking beverages above 65°C (149°F) as probably carcinogenic to humans in 2016. This is a hazard classification based on the available human and animal evidence. It does not mean 64°C is proven safe or that a single sip at 66°C causes cancer. The 2026 UK study may capture differences within temperatures people commonly encounter, but it cannot locate those differences on a thermometer.
An Iranian tea cohort gives a more concrete comparison. In the Golestan temperature study (https://pubmed.ncbi.nlm.nih.gov/30891750/), researchers measured tea temperature at enrollment using a standardized method and asked about usual preferences and waiting time. The 50,045 participants were aged 40 to 75 and followed for a median 10.1 years. During follow-up, 317 developed esophageal squamous cell cancer.
People whose measured tea drinking temperature was at least 60°C had a 41% higher relative risk than those below 60°C, after adjustment for measured factors. People who said they liked tea very hot also had higher risk than those preferring it cold or lukewarm. Waiting less than two minutes after pouring, compared with at least six minutes, was associated with higher risk too. These are three related ways of describing heat exposure; they do not show that everyone should wait exactly six minutes. A large mug, a small cup, a lid, and added milk cool at different rates.
Measured temperature is the Golestan study's strength. Its limitation is context: it involved tea drinking in a region with a high rate of this cancer and different patterns of drink volume and other exposures from the UK. The measured result supports concern about temperature itself, while the much larger UK study shows an association in a Western setting. Neither demonstrates how much an individual person's risk changes after deliberately cooling drinks.
Why an older UK study is not a second independent test
A 2025 UK Biobank hot-beverage study (https://pubmed.ncbi.nlm.nih.gov/39972189/) reported a similar pattern in 454,796 UK Biobank participants. It followed 242 squamous cell cases and 710 adenocarcinoma cases. Very hot preference and greater intake were linked to higher squamous cell risk, with no clear adenocarcinoma pattern. That is useful background, but the 2026 analysis includes UK Biobank again. Counting the 2025 and 2026 UK papers as two independent replications would count many of the same people twice.
The independent temperature measurement comes from the Iranian cohort. Together, the studies make thermal exposure a plausible concern. They do not isolate every route by which heat might affect tissue. The proposed explanation is repeated injury to the lining of the esophagus, but the UK researchers measured cancer diagnoses and drink preferences, not microscopic injury over time. A plausible mechanism should not be presented as an observed chain from a particular sip to a tumor.
There is also a limit to claims about tea or coffee ingredients. The UK association appeared across drink types, which points toward heat, yet the study was not designed to compare every brewing method or ingredient. A person can keep the beverage they enjoy and change how hot it is when swallowed. Our black tea overview discusses the drink more generally; the temperature question here stands on its own research.
A practical way to use the evidence
Make the temperature decision at the first sip, not when the kettle stops. Let a freshly prepared drink stand, stir it, or add cooler liquid if that suits the recipe. Sip when it feels comfortably warm in your mouth rather than painfully hot. Avoid treating a fixed waiting time as a medical rule, because the studies did not test a universal cooling interval for homes, cups, and serving sizes.
The habit is easiest to notice when drinks are repeatedly served at the same high temperature: a thermos, a café takeaway cup, or several refills of tea. If you often burn your tongue or need to interrupt each sip, a cooler serving is a reasonable change. A thermometer is optional; the UK study did not give a validated safe target for every person. The 60°C Iranian comparison and the 65°C agency classification are evidence landmarks, not a personal guarantee on either side of a line.
Temperature is only one part of risk. Avoiding tobacco and limiting alcohol matter for squamous cell cancer, according to the National Cancer Institute's risk and prevention information. Cooling tea should not be treated as a way to cancel those exposures. If swallowing becomes painful or difficult, food sticks, or unexplained weight loss develops, seek medical assessment rather than trying to change drink temperature alone. The Institute's symptom guidance describes these signs and notes that other conditions can cause them too.
The useful conclusion is modest: repeated very hot drinking is associated with higher risk of squamous cell cancer in more than one population, and measured temperature strengthens the evidence. Cooling a drink until it is comfortable is a simple way to reduce the exposure being studied. The exact amount of risk an individual avoids remains unknown.
