Jianhui He1, Sawitri Assanangkornchai2, Le Cai3, Edward McNeil1. 1. Department of Health Economics and Health Management, School of Public Health, Kunming Medical University, Chenggong New City, Kunming, Yunnan Province, P.R. China. 2. Epidemiology Unit, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla 90110, Thailand savitree.a@psu.ac.th. 3. School of Public Health, Kunming Medical University, Chenggong New City, Kunming, Yunnan Province, P.R. China.
Abstract
AIMS: Although varied methods are used in estimating alcohol consumption, there has been no methodological comparison of these methods in China. The aim of this study is to compare different methods used for estimating drinking indices, such as annual per capita consumption, and proportions of risk drinking. METHODS: A cross-sectional household survey using a multistage systematic sampling technique was conducted in Yunnan Province of China. The beverage-specific quantity frequency (BSQF), tri-level, last 7-day (L7D) and beverage-specific yesterday (BSY) methods were used. RESULTS: A total of 977 subjects aged 12-35 years were surveyed. The BSQF yielded the highest annual per capita consumption (2094.2 g), followed by the tri-level (1430.8), L7D (983.0) and BSY (409.7 g) methods. The annual per drinker consumption quantities were 1762.1, 938.5, 865.7 and 493.3 g for tri-level, BSQF, L7D and BSY methods, while the proportions of high-risk drinking were 13.4, 9.4, 4.8 and 2.5%, respectively. CONCLUSION: The BSQF method is more suitable for measurement of annual per capita alcohol consumption in this population. The tri-level approach is a promising substitution of the graduated quantity frequency (GF) method if implemented with caution for both quantity and pattern of alcohol consumption.
AIMS: Although varied methods are used in estimating alcohol consumption, there has been no methodological comparison of these methods in China. The aim of this study is to compare different methods used for estimating drinking indices, such as annual per capita consumption, and proportions of risk drinking. METHODS: A cross-sectional household survey using a multistage systematic sampling technique was conducted in Yunnan Province of China. The beverage-specific quantity frequency (BSQF), tri-level, last 7-day (L7D) and beverage-specific yesterday (BSY) methods were used. RESULTS: A total of 977 subjects aged 12-35 years were surveyed. The BSQF yielded the highest annual per capita consumption (2094.2 g), followed by the tri-level (1430.8), L7D (983.0) and BSY (409.7 g) methods. The annual per drinker consumption quantities were 1762.1, 938.5, 865.7 and 493.3 g for tri-level, BSQF, L7D and BSY methods, while the proportions of high-risk drinking were 13.4, 9.4, 4.8 and 2.5%, respectively. CONCLUSION: The BSQF method is more suitable for measurement of annual per capita alcohol consumption in this population. The tri-level approach is a promising substitution of the graduated quantity frequency (GF) method if implemented with caution for both quantity and pattern of alcohol consumption.