Understanding alcohol consumption trends, especially among heavy drinking groups, is undoubtedly central to alcohol policy debates across the UK, but unveiling the complexity and nature of alcohol use across society is no mean feat. Recent research, however, provides new insights into what and who lies behind recent consumption shifts and how to interpret these in the context of 'drinking types' and 'cultures'.
A new analysis of ‘heavier drinkers’ in Great Britain from 1978-2010, funded by ESRC and published earlier this year, urges against relying on headline consumption data when considering alcohol problems and policy. Rather, we need to recognise the 'rich tapestry' of alcohol consumption: the patterns of use and culture that are spread not only across the whole population but within a range of drinking 'sub-groups'. The research explored 'typologies and dynamics' of heavier drinkers, identifying four 'stable clusters' during the period, with each group showing characteristics that were distinct from both the other heavy drinking groups and the general population.
Between 1978 and 2010 consumption mainly rose, with what has been described elsewhere as 'peak booze' being reached around 2004. This increase in consumption was driven in part by higher levels of wine drinking amongst women, including 'baby boomers' born in the 1940s and 50s. Over this period wine has been increasingly 'democratised': drunk in greater quantities not only by women but across a wider range of income groups. Wine sales increased by 184% between 1980 and 2007 in the context of increasing affordability and availability through supermarkets and the off-trade. A later generation of women, reaching early adulthood in the 1990's, also significantly contributed to rising consumption. However whilst the older 'baby boomer' generation of women fall firmly within the 'wine and spirit cluster' (the only female dominated drinking group), women's drink choices have more recently diversified alongside growing choice in the market.
Importantly, the authors suggest that shifts in overall levels of consumption tend not to occur as a result of new distinct groups or styles of drinking, but rather develop within existing drinking cultures. For instance, while the increase in heavy drinking by younger women during the 1990's and early 2000's was frequently portrayed in the media as 'ladette' culture, this overlooked the still heavier rates of drinking by men whose consumption styles remained largely unchanged. Equally, the much talked about invention of 'alcopops' received disproportionate attention in relation to overall rising consumption, which was more substantially driven by increased home drinking across all drink types, particularly women's wine consumption.
The study’s lead author, Dr Robin Purshouse, said:
“Over the last 30 years of social change, the styles of drinking adopted by heavier alcohol users have stayed remarkably consistent. The rise in heavier drinking over the 1990s in the lead up to ‘peak booze’ was driven by increasing numbers of women and older people adopting these styles. Our findings emphasise the importance of cultural factors, such as drinking styles, as key components in the policy debate surrounding heavy alcohol use.”
The study overlaps with research exploring drinking cultures published last year (also led by members of the Sheffield Alcohol Research Group), which also described a more complex picture than that often portrayed in the media. That study found nearly half (46 per cent) of all drinking occasions 'involved moderate, relaxed drinking in the home', potentially indicating the context in which the majority of 'low risk drinkers' do so. Certainly there is relevance to further understanding low risk drinking nuances too, especially when considering how heavier drinkers tend to describe their drinking as in line with the norm. However the SARG findings also support evidence of 'pre-loading', and 10 per cent of all drinking occasions involved groups of friends moving between home and pub drinking, consuming the average weekly recommended guideline of 14 units on one occasion. Other identified patterns included drinking at home alone (14 per cent of occasions), light drinking at home with family (13 per cent), light drinking at home with a partner (20 per cent) and heavy drinking at home with a partner (nine per cent). See here for Conversation article by SARG researcher John Holmes.
And new insight into under-estimation in national survey data
A recent study has explored to what extent national consumption data based on surveys may under-represent heavier drinkers due to 'non-response bias'. It has often been shown that heavier drinkers may be less likely to be respond, or may require more extensive efforts to recruit to surveys. The study analysed how the number of contact attempts to reach participants varied by drinking status and socio-demographic characteristics, identifying evidence for a significant 'non-response bias’ among heavier drinkers. When modelled, it was estimated that accounting for non-response bias may lead to a 12.6% increase in men’s weekly drinking and 20.5% in women.
Lead author Dr Sadie Boniface said:
"Our recent paper presents one way of looking at the likely impact of non-response bias on survey measures of alcohol consumption, similar to previous studies in New Zealand and Canada. Our findings agree with other studies, finding that people with a more hazardous or harmful drinking pattern are harder to reach for surveys. This study underlines the importance of extended efforts to recruit and follow-up participants in research studies in order to reduce the impact of this bias."
Overall, recent findings such as these remind us that while the overall level of consumption across the population is an important indicator of trends, within any society drinking behaviours are varied and diverse. Harmful patterns may emerge among one group, or within one set of drinking environments, while trends may improve elsewhere.
Understanding the contexts and multiple factors at play is therefore important in developing and refining responses to alcohol harm. Individually targeted interventions may be generally considered effective, but still face significant delivery challenges within the current limitations of research, policy and practice. Indeed the paper states that the need for more nuanced understandings and responses does not itself call into question models that suggest a 'structural relationship between overall population consumption levels and harmful consumption'. Indeed there are many complex questions facing the many levels of alcohol policy and interventions - see here for selected events in 2017 that will be attempting to further answer some of these.
With thanks to James Nicholls, Alcohol Research UK, for contributions.