Potential link between tastes preference and digestive system cancer hospitalisations in Fujian Province, China: big data analytics.

[OBJECTIVE] The study aimed to utilise internet big data to quantify the taste preferences of residents in Fujian Province and to explore the relationship between dietary taste preferences and hospitalisation rates for digestive system cancers.

[DESIGN] The study employed an associative design using internet big data to analyse dietary behaviour and its association with hospitalisation rates for digestive system cancers. GeoDetector methods were used to compare the association between rural residents' hospitalisation rates and their taste preferences.

[SETTING] This study utilised internet recipe data to collect cuisines taste information. By integrating this with categorised restaurant data from point of interest sources across various regions in Fujian province, it quantitatively analysed the regional taste preferences of people.

[PARTICIPANTS] Data from seventy-two counties in Fujian cover most of the province. Included 154 686 hospitalisation records for digestive system cancers (2010-2016) from the New Rural Cooperative Medical Scheme database, 16 363 recipes from Internet and data from 30 984 restaurants through Amap.

[RESULTS] The study found pungent to be the prevalent taste in Fujian, with salty, spicy and sour following. Coastal areas favoured stronger tastes. Spatial analysis showed taste preferences clustered geographically, with Sour and Fat tastes having an association with liver and colorectal cancer (CC) hospitalisations, though with modest association values (0·110-0·199).

[CONCLUSIONS] The study found significant spatial clustering of taste preferences in Fujian Province and an association between Sour and Fat tastes preference and hospitalisation rates for liver and CC, suggesting a dietary taste-cancer link.