Very hot drinks may damage the food pipe and increase cancer risk
This is essentially competent, good-faith health journalism reporting a real, peer-reviewed Oxford study on the day of publication — not a polemical or agenda-driven piece. Its weaknesses are structural to the genre rather than ideological: it leads with a relative risk figure (threefold) before grounding the reader in the low absolute baseline; it conflates oesophageal squamous cell carcinoma with oesophageal cancer generally; it never mentions that drink temperatures were self-reported rather than measured; and it quotes only voices from the study's own funder (Cancer Research UK) without independent critical commentary. The IARC had already classified very hot beverages as probably carcinogenic in 2016, so this study is confirmatory for a Western population — not the novel link-establishment the article implies. Readers should note that a 1% lifetime risk tripled is still a 3% lifetime risk, that confounding in observational dietary studies is substantial, and that the practical implication (let your tea cool a bit) is already the norm for most UK drinkers. The article is worth reading but should be paired with the original paper's methods section before drawing strong behavioural conclusions.
Summary
Likely motivation
What this article didn't consider
The article's implicit thesis is that very hot drinks represent a meaningful, modifiable cancer risk worth changing behaviour over. The strongest opposing case: the study is observational and relies on self-reported temperature preference, not thermometer measurements, so the 'very hot' category is subjective and imprecise. The threefold relative risk sounds alarming but translates to an absolute increase well under 1 percentage point for a cancer that kills around 2,500 people a year in the UK; behaviour-change messaging focused on drink temperature risks distracting from the vastly larger, causally better-established risks of smoking, alcohol, obesity, and Barrett's oesophagus. Epidemiological associations at this scale of confounding — where 'very hot drink' drinkers likely differ from 'warm drink' drinkers in dozens of dietary and lifestyle ways — routinely fail to replicate as causal relationships.
- The IARC classified very hot beverages (above 65°C) as 'probably carcinogenic' in 2016, a full decade before this study — the article implies novelty ('the largest to date') without noting this classification was already in place and already influenced guidance in many countries.
- The study cannot measure actual drink temperatures: participants self-reported preference ('very hot', 'hot', 'warm'). People who describe their tea as 'very hot' are not proven to drink at 65°C; this is a critical methodological limitation that the article omits entirely.
- The article covers only oesophageal squamous cell carcinoma (SCC) but does not make this distinction clearly enough: hot drinks showed no increased risk for oesophageal adenocarcinoma (AC), the more common subtype in the UK. The article's framing implies a general oesophageal cancer link when the finding is subtype-specific.
- Comparable populations: studies in Iran, China, and East Africa — where drinking scalding tea is deeply embedded — show much stronger absolute rates of ESCC than the UK, suggesting the UK's typical drink temperature may already be below the truly dangerous threshold. The article mentions South America but doesn't draw this comparative calibration.
- The article does not report confidence intervals for the key threefold risk estimate. The actual study shows a hazard ratio of roughly 3.0 with wide confidence intervals reflecting a small number of ESCC cases even in a near-million-person cohort — the precision of the estimate is much lower than the headline figure suggests.
- No mention is made of the confounding problem inherent in self-reported dietary studies: people who habitually drink very hot beverages may also differ in other dietary habits (e.g., lower fruit and vegetable intake, higher rates of smoking in certain cultural subgroups) that drive ESCC risk independently.
- The '1 in 10 oesophageal cancers could be avoided' claim is a population attributable fraction estimate that depends heavily on how many people actually drink at genuinely dangerous temperatures — a number the study cannot establish because temperatures were self-reported.
Cancer Research UK both funded the study and is quoted twice in the article providing reassurance and context. This is not sinister, but it means the funder is also the article's main independent validator — a circularity the article does not disclose. The charity benefits from public attention to cancer risk factors as it supports their fundraising and public-health profile; the BBC benefits from a high-traffic health-risk story timed to a journal publication. No political or ideological strategic framing is detectable here beyond the standard science-journalism incentive to present findings as more novel and actionable than they may be.
Logical fallacies
- Relative Risk Inflation / Base Rate Neglect
“people who had their drinks 'very hot' had around a three times higher risk of the cancer than 'warm' tea or coffee drinkers”
Foregrounding a threefold relative risk without first anchoring the reader to the absolute base rate (lifetime risk ~1%) is a classic base-rate neglect pattern in health journalism. A relative risk multiplier is only meaningful after the baseline probability is established. The article does eventually mention low overall risk, but the structure buries absolute risk behind the alarming relative figure.
- Conflation of Correlation with Causation
“Regularly drinking beverages like tea and coffee at very hot temperatures increases the risk of cancer”
The study is observational (prospective cohort with self-reported exposure). The lead sentence uses 'increases the risk' in causal language; the study can only establish association. The observational design — which the article describes accurately later — cannot rule out residual confounding.
- False Precision
“around one or two in every 10 oesophageal cancers could be avoided if people stopped drinking very hot beverages”
The population attributable fraction stated here depends on knowing both the true prevalence of 'very hot' drinking at genuinely dangerous temperatures and the causal effect size — neither of which the study can establish given self-reported temperatures and observational design. Presenting this as a concrete preventive fraction overstates what the data support.
Bias indicators
- Novelty Bias / Omission of Prior Classification
“make this the largest to date to establish the link”
The WHO's IARC had already classified very hot beverages (above 65°C) as 'probably carcinogenic' in 2016 based on prior international evidence. The framing as a landmark new establishment of the link elides a decade of prior scientific and regulatory consensus.
- Subtype Conflation
“increases the risk of cancer of the food pipe or gullet, known as oesophageal cancer”
The finding applies specifically to oesophageal squamous cell carcinoma (SCC), not to oesophageal adenocarcinoma (AC) — the more common subtype in the UK. The article uses the general term 'oesophageal cancer' in the lede without this crucial distinction, which inflates the apparent breadth of the risk.
- Funder Circularity
“Fiona Osgun from Cancer Research UK, which funded the study”
The article's two 'independent' expert voices are both from Cancer Research UK, which funded the study. This is disclosed in a subordinate clause but not flagged as a potential conflict; it means there is no truly independent commentary in the piece.
- Measurement Limitation Omission
“defines any drink at 65 degrees Celsius or above as 'very hot' and posing a potential risk”
The article never mentions that drink temperatures were self-reported by preference category, not measured. Readers are left believing 65°C is a validated threshold established by direct measurement in this study, when it is actually an extrapolated IARC benchmark applied to subjective self-report data.
Loaded language
Missing context
- The study relied on self-reported preferred temperature categories, not thermometer measurements of actual drink temperatures — a major methodological caveat absent from the article.
- The threefold relative risk applies only to oesophageal squamous cell carcinoma (SCC), not to adenocarcinoma (AC), the more prevalent oesophageal subtype in the UK — a critical distinction blurred in the lede.
- The IARC already classified very hot beverages (>65°C) as 'probably carcinogenic' in 2016, making this study confirmatory for a Western population, not a novel discovery of the link.
- No confidence intervals are given for the hazard ratio, obscuring the statistical precision (and uncertainty) of the threefold estimate.
- The number of SCC cases in the study was small (around 242–600 depending on cohort), meaning the relative risk estimate has wide uncertainty despite the large overall sample.
- Potential confounders — e.g., that very hot drink preference may correlate with socioeconomic status, smoking culture, or dietary patterns that independently drive ESCC risk — are not discussed.
- The article does not note that typical UK tea-drinking temperatures (around 60–65°C once milk is added and a minute or two has passed) may already fall below the 65°C danger threshold for most people, meaning the practical risk for the average UK tea drinker may be lower than implied.
- Both quoted experts (Cancer Research UK) represent the study's funder, providing no independent critical commentary.
Author & publication
- Author
- Michelle Roberts
- Publication
- BBC News
- Funding notes
- BBC News is primarily funded by the UK television licence fee, a statutory charge set by the UK government. It has no advertising revenue for domestic output and no private ownership. Cancer Research UK, which funded the underlying study, is also quoted in the article.
- Track record
- Roberts is BBC News's Digital Health Editor and a qualified doctor. Her body of work spans standard public-health journalism (statins, weight-loss drugs, vaccines, cancer risk factors) and follows the BBC's house style of leading with peer-reviewed study findings. No evidence of ideological or commercial affiliations beyond her BBC role was found.