Abstract
Background: In 2016, the International Agency for Research on Cancer Monographs programme classified drinking very hot beverages as a probable carcinogen in the oesophagus based on limited epidemiological evidence, mainly from South America, but also from a few studies in Asia and Africa. Since then, individual case-control studies in Africa and Asia, and cohort studies from Asia and Europe have reported positive associations between drinking very hot beverages or eating very hot foods and increased risk of oesophageal cancer (EC), predominantly squamous cell carcinoma (ESCC). 
Objectives: To examine the association of consuming hot or very hot beverages or foods and oesophageal cancer risk in a pooled case-control studies in Africa and updated meta-analysis of existing case-control and cohort studies conducted published from 2015-present.
Methods: We pooled 7 hospital-based case-control studies from the African oesophageal cancer consortium-aetiology pooling project (AfrECC-EPP) involving clinically, endoscopically or histologically confirmed ESCC cases and controls. The temperature of drinking tea, coffee, porridge or soup or eating food was self-reported by participants and harmonised into non-drinkers, cold/warm, hot, and very hot/extremely hot. Odds ratios for hot and very hot/extremely hot vs. non-drinkers and cold/warm drinkers were computed generalised linear mixed models adjusted for traditional ESCC risk factors. Pooled ORs were included in a preliminary random effects meta-analysis of 15 case-control studies and 8 cohort studies, with a total of 8,263 EC cases and 963,733 controls.
Results: In the AfrECC-EPP, 2,439 oesophageal squamous cell carcinoma (ESCC) cases and 2,631controls were pooled from Kenya, Tanzania, Malawi and Zambia. The mean (standard deviation) age of cases and controls were 58 (14) and 56 (14) respectively, and 65% of cases were men. After adjusting for age, sex, education, drinking alcohol, and smoking, Very/extremely hot temperature was associated with increased risk of ESCC (odds ratio [95% confidence intervals, 1.39 [1.16-1.67]), in men (1.31 [1.03-1.65]) and women (1.54 [1.16-2.05). These positive associations were consistent in female non-alcohol drinkers and non-smokers and were positive but slightly weaker in male alcohol drinkers or smokers. Preliminary results from the meta-analysis showed an increased risk of EC and ESCC with hot and very hot beverage or food consumption.
Conclusions/Implications for practice and policy:   Consumption of very hot food or drinks was associated with total EC and  ESCC, and potentially explains excess ESCC incidence in female non-alcohol drinkers and non-smokers in Africa’s EC corridor.

Keywords: Esophageal squamous cell carcinoma, hot tea, thermal injury, meta-analysis, pooled analysis