ABSTRACT

Studies in Asia, Africa, South America and the Middle East have consistently shown that tea or mate drunk at very high temperatures (~70°C) increases oesophageal cancer risk, but evidence is limited for drink temperatures typical of Western populations. In analyses of two large UK cohorts (977,282 participants), we used Cox regression to estimate hazard ratios (hereafter referred to as RRs) and 95% confidence intervals (CIs) for oesophageal cancer subtypes by self-reported preferred drink temperature, our primary focus and frequency of hot drink (tea and coffee) consumption. Pooled estimates were calculated using inverse-variance-weighted averages of cohort-specific log RRs. During an average follow-up of 14.2 years (Million Women Study [MWS]) and 11.1 years (United Kingdom Biobank [UKB]), 2348 incident oesophageal cancers occurred (1045 squamous cell carcinoma [SCC], 1303 adenocarcinoma [AC]). Consumption of drinks at a 'very hot' versus 'warm' temperature was associated with a 3-fold greater SCC risk (RR = 3.17 [95% CI: 2.49, 4.03]) but not with AC risk (0.92 [0.76, 1.12]), after adjustment for consumption frequency. After controlling for drink temperature, daily consumption of 6+ versus < 6 hot drinks was associated with a 71% increase in SCC risk (1.71 [1.51, 1.94]), with little association with AC risk (1.19 [1.06, 1.33]). In UK adults, an estimated 14% of SCC oesophageal cancers could be avoided if individuals who reported consuming their drinks at a 'very hot' temperature instead consumed them at a temperature more representative of the temperature range described here as 'hot'. The fact that we observed a clear trend in risk across self-reported temperature categories suggests that further limiting the temperature at which individuals consume their drinks to something well below average for this population may well avoid an even greater proportion of SCC cases. This large UK study of middle-aged adults, amongst whom hot drinks are widely consumed, provides clear evidence that temperature and frequency of hot drink consumption represent useful targets for oesophageal SCC prevention.

‍