The distribution of alcohol consumption and heavy episodic drinking across British drinking occasions in 2019: a cross-sectional, latent, class analysis of event-level drinking diary data

Alcohol is a major cause of poor health but few studies have examined the contexts in which people consume alcohol. We aimed to develop a typology of British alcohol drinking occasions in 2019, and estimate how the population's alcohol consumption and heavy drinking (>6 units for women, >8 units for men) is distributed across occasion types.The typology is based on a cross-sectional latent class analysis of data from drinking diaries collected via quota samples of adult drinkers (≥18 years) resident in Great Britain. Three latent class models use these characteristics as input variables to estimate typologies for off-trade occasions (eg, home-drinking), on-trade occasions (eg, pub-drinking), or mixed-trade occasions (eg, home-drinking then pub-drinking). We calculated the mean units of alcohol consumed per occasion, and the proportion of observed consumption and of observed heavy-drinking occasions within each type.A total of 17 821 diaries reported characteristics of 43 089 drinking occasions in 2019, including the location, timing, companions, motivations, accompanying activities, and alcohol consumed. The analysis identified 15 occasion types: four off-trade only, eight on-trade only, and three mixed-trade. Mean consumption per occasion varied between 4·4 units in the Family Meals type and 17·7 units in the Big Nights Out With Pre-loading type. Mean consumption exceeded 10 units in all mixed-trade occasion types and in the Off-trade Get Togethers (10·4 units), Big Nights Out (11·1 units), and Male Friends At The Pub (10·2 units) types. In 13 occasion types, more than 25% of occasions involved heavy drinking. Conversely, 41·7% of Big Nights Out and 16·4% of Big Nights Out With Pre-loading were not heavy drinking occasions.Heavy drinking is common in most types of British drinking occasion and particularly within mixed-trade drinking occasions. However, lower consumption levels are also common in types often characterised as heavy drinking practices. These results can support the design of alcohol policies to ensure they target the most harmful drinking practices. Although this study provides, to our knowledge, the first detailed analyses of the distribution of British alcohol consumption and heavy drinking occasions across drinking contexts, its generalisability is limited by the non-random sampling method.Economic and Social Research Council Grant Number ES/R005257/1.