The impact of promoting revised UK low-risk drinking guidelines on alcohol consumption: interrupted time series analysis

Article history

The research reported in this issue of the journal was funded by the PHR programme as project number 15/63/01. The contractual start date was in November 2015. The final report began editorial review in May 2019 and was accepted for publication in January 2020. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PHR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

© Queen’s Printer and Controller of HMSO 2020. This work was produced by Holmes et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2020 Queen’s Printer and Controller of HMSO

Guidelines for low-risk drinking

Drinking guidelines have proliferated internationally since the 1980s and health authorities in most high-income countries now issue guidelines for alcohol consumption. 8 These usually recommend a quantified level of alcohol consumption that drinkers should not exceed over a specific time period (e.g. per day or per week). The consumption level is usually expressed in standard drinks or units of alcohol. In the UK, 1 unit is 8 g or 10 ml of pure ethanol and is approximately half a pint of beer, 75 ml of wine or one 25-ml shot of spirits, depending on the alcoholic strength of the drink. The guideline consumption threshold is also usually accompanied by textual information that may include, among other things, the source of the guideline (e.g. the CMO recommends . . .), a reason to listen to the advice (e.g. to reduce your risk of alcohol-related disease . . .) and supplementary recommendations (e.g. do not drink every day).

Figure 1 shows that drinking guidelines often vary markedly between countries. There may be a number of reasons for this. First, the purpose of guidelines often differs between countries. For example, Denmark provided guidelines to help people avoid high-risk levels of alcohol consumption until 2010, whereas the UK has tended to issue guidelines to promote low-risk consumption. 8 Other countries use guidelines primarily to inform government policy (e.g. to assess levels of excessive drinking) or clinical practice, rather than as health promotion information for the public. Second, the riskiness of drinking may vary between countries depending on the prevalence of competing or compounding risk factors, such as smoking and obesity. Third, guidelines are usually expressed in units or standard drinks, but these measures vary widely across countries. For example, 1 unit is 8 g of pure ethanol in the UK, 10 g in Ireland, 12 g in Finland, 14 g in the USA and 20 g in Austria. 9 If guidelines are to be easily understood by the public and not expressed in decimals, variation in the definition of 1 unit necessitates variation in drinking guidelines.

FIGURE 1.

Weekly drinking guidelines for women and men in selected countries in 2016. Shaded countries at the bottom of the figure are those with the same guideline for men and women.

Guidelines also vary over time. The earliest quantified guideline dates back to Anstie’s limit of 28–43 g per day, which was developed in the 1860s and used in life assurance in the UK and USA. 10 More recently, UK health authorities published ‘safe limits’ for men of 64 g per day in 1979 and then 144 g per week in 1984, followed by ‘sensible limits’ of 168 g per week in 1987 and then 24–32 g per day in 1995. 11 These changes reflect improvements in epidemiological evidence, societal standards regarding the acceptability of health risks and the purpose of guidelines, as described above. They also reflect changes in how guidelines are set. As described by Holmes et al. ,12 the evidence used by guideline development committees is increasingly likely to include bespoke epidemiological modelling exercises in addition to reviews of scientific literature.

Despite their widespread use, drinking guidelines do not enjoy universal support from public health actors. As discussed below, their effectiveness as a public health intervention is questioned,13,14 and some researchers believe that they are counterproductive. For example, Casswell15 argued in 1996 that guidelines are unlikely to affect behaviour when presented in the context of the pro-alcohol messages that suffuse contemporary media, a view shared by the authors of the leading summary of alcohol policy effectiveness. 13 In 2012, Casswell16 reiterated this point and added that promoting guidelines draws government away from more effective interventions. Similarly, Bellis17 argues that, when implemented in practice, drinking guidelines serve the interests of the alcohol industry as, ‘rather than explaining risks, [they] can read more like an alcohol promotion slogan’ that presents recommended consumption levels.

Other public health actors view guidelines more positively. Heather18 argues that they are essential as the public has a right to know about alcohol-related health risks and they facilitate discussions about alcohol consumption in clinical practice. Marteau19 further suggests that, although guidelines may not lead directly to reductions in alcohol consumption, they increase public concern about the risks of alcohol consumption and may make it easier to introduce effective alcohol control policies.

Drinking guidelines in the UK, 1995–2016

Prior to 2016, the UK drinking guidelines had not changed since 1995, when the Sensible Drinking report20 recommended that men should not regularly consume more than 3–4 units a day and women should not regularly consume more than 2–3 units a day, where regularly means drinking this amount every day or on most days.

The House of Commons Science and Technology Committee launched an inquiry in 2011 to assess whether or not the 1995 guidelines were ‘robust’ as a foundation for alcohol policy. 11 It sought submissions on the evidence underpinning the guidelines, whether or not that evidence could be improved, how well the government communicates the guidelines and how the guidelines compare with other countries. The report concluded that the passage of time, greater evidence on the causal relationship between alcohol and cancer and increased scientific scepticism regarding the health benefits of moderate drinking meant that the guidelines should be reviewed by an expert group.

In 2012, the UK Government published its alcohol strategy5 and announced that the CMO would lead a review of the guidelines, which ran from 2013 to 2016. Although the review was led jointly by the four UK CMOs, it was conducted primarily by a guideline development group (GDG) that mainly comprises civil servants, clinicians and academics with expertise in health and behavioural science. The GDG considered published studies, commissioned new systematic reviews, epidemiological modelling and focus group research, and heard presentations from expert advisors. In 2014, its interim report concluded that that there was sufficient new evidence to justify changes to the guidelines. The CMOs accepted this conclusion.

In 2016, the GDG published its final report21 along with the new draft guidelines,22 which recommended that the guidelines should be revised to state that men and women could keep their risks from alcohol consumption low by consuming no more than 14 units per week. Additional accompanying guidance recommended that drinkers consuming ≥ 14 units per week spread their alcohol consumption across several days and have at least 2 drink-free days per week. These recommendations differed from the 1995 guidelines in three main ways. First, they switched from a daily to a weekly guideline, citing evidence that the previous guideline was often misinterpreted as a limit on consumption on any single occasion and seen as less useful as a result. 23 Second, they set the same guideline for men and women, a decision that triggered substantial comment and criticism as only two other countries take this approach (see Figure 1). Third, they reduced the previous guideline for men. That guideline was often interpreted as ≤ 14 units per week for women and ≤ 21 units for men, a legacy of pre-1995 guidance, and the revisions were, therefore, viewed as reducing the guideline for men by one-third.

The GDG also revised the guidance in two other key areas. First, it updated the guidance on drinking alcohol during pregnancy. Previously, the guidance in Scotland was that women should not drink during pregnancy, whereas the guidance in England was that women should not drink more than 1 or 2 units once or twice per week. The revised guidance removed this inconsistency and took a precautionary approach that recommended:

. . . the safest approach is not to drink alcohol at all, to keep risks to your baby to a minimum.

Reproduced with permission. 24 © Crown copyright 2016. Contains public sector information licensed under the Open Government Licence v3.0

Second, the GDG provided updated guidance on single occasions of drinking and elected not to provide a specific guideline but instead recommended that drinkers keep their risk to a low level by:

Limiting the total amount of alcohol you drink on any single occasion, drinking more slowly, drinking with food and alternating with water, and planning ahead to avoid problems.

Reproduced with permission. 24 © Crown copyright 2016. Contains public sector information licensed under the Open Government Licence v3.0

Notably, this guidance did not incorporate the widely used definition of binge drinking as the consumption on a single occasion of more than 6 units (women) or more than 8 units (men). The GDG elected not to set a guideline consumption threshold for single drinking occasions as risks strongly vary substantially between individuals depending on their ability to reduce such risks by other means. 21

The new guidelines were announced publicly by the CMOs in January 2016 and were formally adopted by the UK Government in August 2016 following a public consultation. 25

The effectiveness of drinking guidelines

Little evidence is available on the effects of drinking guidelines, and reviews note that there have been no rigorous evaluations of the impact of producing, revising or promoting drinking guidelines on alcohol consumption or on other factors related to behaviour change. 13,26–28 This is somewhat surprising as public guidance on a wide range of health-related behaviours, including smoking, physical activity, drink-driving and nutrition, has produced small to moderate effects on knowledge, attitudes and behaviour. 28–32

A number of factors may explain the lack of evidence on the effectiveness of drinking guidelines. Opportunities to evaluate the impact of guidelines are rare as revisions take place at intervals of ≥ 10 years and, when they do happen, the new guidelines are often not promoted via large-scale campaigns after the initial announcement. Furthermore, few countries routinely collect data that would facilitate robust evaluation. For example, data on awareness, knowledge or exposure to guidelines are not a common feature of public health surveys, and alcohol consumption data are typically collected on an annual basis only, which makes robust analysis of intervention effects difficult as they are likely to be modest given the scale of promotional activity. For these reasons, the small literature that does attempt to examine effects of promoting drinking guidelines relies on relatively weak research designs.

Studies in several countries have used cross-sectional analyses to examine public knowledge of guidelines. These analyses generally find that large minorities of the population are aware of drinking guidelines and can correctly identify guideline consumption levels, but there is little evidence that this influences drinking behaviours. 33–35 For example, studies in Australia, Denmark and the UK have examined changes in guideline-related knowledge and perceptions over time using annual cross-sectional data. 36–39 In each case, the results suggest that knowledge of guidelines improved over time. In Denmark, this appeared to be linked to annual mass media campaigns to promote the guidelines, but, in Australia, beliefs about what constitutes safe drinking also changed.

The UK data are particularly relevant to the present study. The Office for National Statistics (ONS) Opinion Survey collected data on knowledge of units and guidelines between 1997 and 2009 and summarised the findings in a 2010 report. 40 The report shows that the proportion of drinkers who had heard of units rose from 79% in 1997 to 90% in 2009. Similarly, the proportion who knew how much beer, wine and spirits constituted 1 unit also rose by up to 10–15 percentage points. In 2009, 63% of respondents knew that 1 unit of beer was approximately half a pint, 69% knew that 1 unit of spirits was a single measure, but only 27% knew that 1 unit of wine was less than a 125-ml glass. The proportion of respondents who had heard of guidelines increased from 54% in 1997 to 75% in 2009. However, only 44% of men and 55% of women in 2009 correctly identified the guideline for their sex, an improvement of only 6 percentage points among men and of 14 percentage points among women over a period of 12 years. Moreover, only 13% of respondents kept a check on their units, with no increase in this proportion between 1997 and 2009. There was little variation in these measures by sex, but heavier drinkers generally showed greater awareness and knowledge of the guidelines and of units.

Prior to the present study, the most detailed evaluation of promoting drinking guidelines focused on an Australian campaign aiming to raise awareness of guidelines and alcohol-related cancer risks. 41 The study used repeat cross-sectional surveys of 150–200 women at three time points over a 15-month period and concluded that multiple waves of advertising led to good recognition and awareness of the campaign content. Promoting drinking guidelines also led to increased cancer risk awareness, knowledge of guidelines and intentions to reduce alcohol consumption among heavier alcohol drinkers in particular. However, there was no evidence of impacts on drinking behaviour.

In addition to these quantitative analyses, a number of qualitative studies have sought to understand how drinkers respond to and use drinking guidelines. These studies report that drinkers typically disregard guidelines as they do not align well with their current drinking practices and rely on drinkers understanding units, which is often not the case. 23,42,43 Instead, drinkers monitor their consumption through embodied means; that is, they make judgements on the riskiness of their drinking based on how intoxicated they feel and their previous experiences of alcohol and its after-effects.

Effectiveness of alcohol health warning labels

Although there is little robust research on the effectiveness of promoting drinking guidelines, a separate body of research exists examining the effectiveness, potential effectiveness and appropriate design of alcohol health warning labels. This literature is important, as approximately 80% of products sold in the UK in 2014 included the CMO’s guidelines on their label,44,45 and this is a key location where drinkers are likely to encounter the guidelines when they are not being promoted via more active means. 46 More generally, information on labels is argued to be a potential influence on social norms around alcohol consumption and increase support for other alcohol policies. 47

The USA introduced mandatory health warning labels on alcoholic products in 1989, which read ‘Government Warning: (1) According to the Surgeon General, women should not drink alcoholic beverages during pregnancy because of the risk of birth defects. (2) Consumption of alcohol impairs your ability to drive a car or operate machinery, and may cause health problems’. 48 Evaluations of the impact of these labels constitute a large proportion of the alcohol labelling literature and the highest-quality evidence on the effectiveness of alcohol health warning labels as a public health intervention. 49 Overall, the evaluations suggest that awareness and recall of USA alcohol warning labels increased initially,48,50,51 but plateaued after around 3–4 years in both adults and adolescents. 52,53 As with drinking guidelines, there was little evidence that these gains in awareness and knowledge translated into changes in alcohol-related behaviours, such as reduced drink-driving or alcohol consumption. 49,53,54

More recent research has used survey, focus group and experimental methods to examine several themes relating to how consumers may use and respond to information and warnings provided on alcohol product labels. For example, focus group research in Australia found that participants favoured labels providing messages related to short-term risks and also found messages with positive rather than fear-based or numerical framings, general health messages rather than ones relating to cancer or specific cancers, and risk-based rather than deterministic messages more believable, convincing and personally relevant. 55 A small experimental study of students at a UK university56 found no significant differences between text-based messages and pictorial messages in terms of fear arousal and intentions to reduce or stop alcohol consumption, and Hobin et al. 57 obtained a similar finding in an experimental study with Canadian adults. This is in contrast to a general view that pictorial messages are more effective,58 although Hobin et al. 57 obtained a similar finding in an experimental study.

The UK’s Royal Society for Public Health surveyed drinkers to determine what information they find most useful. 47 The results revealed that drinkers view alcoholic strength and units as the most important information to include on labels and consider warnings against drink-driving and drinking during pregnancy as moderately important; however, drinking guidelines and links to online information are viewed as substantially less important. 47 Thomson et al. 59 found that drinkers in Australia supported the provision of health information and warnings on product labels, whereas other studies found that drinkers in Australia report poor recall or use of industry-backed logos and weblinks on product labels60 but good awareness of standard drink labels. 61 In Canada, drinkers demonstrate a high level of ability to use standard drink information on product labels, and a focus group study of drinkers in Yukon, Canada, also found that detailed product labels including a health warning, guidelines and standard drink information were seen as of intrinsic value. 57,62 The information provided in this study was viewed by participants as new, useful, important and something they had a right to know. Participants also suggested that it could prompt new conversations and potentially influence behaviour. However, other studies suggest that, particularly among some younger drinkers, standard drink information may be used to maximise the volume of alcohol purchased rather than to moderate consumption. 47,61,63 More broadly, a criticism common across research on alcohol labels is that the information provided on real-world labels is too small, discreet, vague or ambiguous, or otherwise lacking in visual impact, to inform or influence consumer behaviour. 44,55,60

The breadth of findings here and the reliance on small-scale, cross-sectional, focus group or experimental research are perhaps the greatest weaknesses of the literature. Hassan and Shiu64 reviewed studies of the efficacy of alcohol warning labels published between 2000 and 2015 and concluded that the literature is ‘a weak and fragmented evidence base to guide policy-makers or practitioners, which requires further research on most key questions’. 64 They particularly criticise the lack of consistent research approaches and measures, which prevents comparison across studies, and the lack of long-term research on the effects of labels. As with the literature on drinking guidelines, we further note that few of these more recent studies following the US warning label evaluations provide evidence on whether such labels can or do affect drinking behaviour. One exception is Maynard et al. ,63 who found that providing calorie or unit information did not affect alcohol consumption by UK undergraduates in a laboratory study. Maynard et al. ’s63 additional qualitative analyses suggest that this is because such information fits poorly with drinking to get drunk or socialise, echoing findings from previous studies. 23,43

Despite their apparent lack of direct impact on alcohol-related behaviours, alcohol health warning labels may, like drinking guidelines, be important in other ways. The US labels are often viewed by the authors of evaluation studies as important because they inform consumers about alcohol-related risks and reinforce this information, even if they do not affect drinking behaviour. 52 Others note that such labels help to stimulate conversations about alcohol and its risks as well as changing social and cultural norms, potentially leading to behavioural changes in future generations,47,54,62 or greater acceptance of alcohol control measures. 19,47 More simply, warning labels may fulfil consumers’ right to know about the risks of a product they are purchasing. 47,62

The literature on the evaluation of alcohol labels has three main implications for our evaluation of drinking guidelines. First, it supports the evidence from the limited literature on the effectiveness of promoting drinking guidelines in showing that health-related information pertaining to alcohol can be recalled and used by drinkers. Second, it also supports the evidence that such information does not generally lead to changes in alcohol consumption behaviour. As suggested by behaviour change theory, such behaviour changes may require well-designed and implemented messaging. 65 In this context, the text around the numerical guideline consumption level may matter as much as the guideline consumption level itself and the content of promotional activity may be as important as its presence. Third, despite the importance of such design and implementation, there is insufficient evidence available to date to identify clear recommendations regarding what is effective in the design of drinking guidelines or guideline-related messaging. Literature from health promotion in other areas, particularly tobacco warning labels, can provide some guidance, but it is unclear whether or not this can be translated unproblematically to alcohol, the risks of which are smaller in scale, different in nature and interact with more positive public attitudes. 64 This means that it is difficult to assess if we would expect the guidelines as written to be effective or what effects they may have.

Rationale for the present study

Despite concerns about their effectiveness, drinking guidelines still play an important role in alcohol policy debate in the UK and internationally. Successive UK governments have been reluctant to increase regulation of the alcohol market and have focused instead on promoting ‘responsible drinking’ with the support of alcohol producers and retailers. 66 The 2012 alcohol strategy5 initially departed from this approach by promising the introduction of minimum unit pricing (MUP) for alcohol, a greater role for public health within alcohol licensing and possible restrictions on price-based promotional deals for alcohol. However, all of these measures were removed from the strategy following a public consultation,67 leaving the review of the drinking guidelines as one of the headline measures alongside increased attempts to identify and deliver brief interventions to drinkers within primary care. The 2016 revisions to the UK drinking guidelines therefore represented both a central plank of government alcohol policy and a rare opportunity to undertake a prospective evaluation of a widely used but under-researched intervention.

Designing an evaluation is not straightforward as promoting drinking guidelines is a complex intervention. Promotional activity is delivery by many different organisations and actors in varying forms, with an extended causal chain from implementation to outcome. Organisations involved include governmental and non-governmental bodies, such as Public Health England (PHE), the NHS, Drinkaware (London, UK) (an industry-funded charity), alcohol producers, retailers and social aspect organisations, public health charities, and news media. Individual actors include social media users, public commentators, individual politicians and the general public to the extent that each of these discusses the guidelines. Likewise, promotional activity ranges across mass media campaigns, interactive social media, consultations with health professionals, product labels, point-of-sale advertising and general public debate. Behaviour change theory suggests that any changes in alcohol consumption that arise from this network of promoters and promotion will be mediated by changes in behavioural influences, including individuals’ knowledge of the guidelines and alcohol-related risks, their motivations and their social context. 65,68 Therefore, a robust and comprehensive evaluation requires an understanding of not just the effects of promoting drinking guidelines on key outcomes, such as alcohol consumption, but also how and by whom guidelines are promoted and the effects of this promotion on proximal outcomes, such as behavioural influences. Appendix 1 presents a causal logic model indicating our understanding of the connections between these processes and informs the design of the evaluation.

Aims and objectives

The aim of this study was therefore to conduct a prospective evaluation of promoting revised UK drinking guidelines on alcohol consumption behaviour. To do so, it sought to meet the following objectives:

1. to document the timing, audience and content of major promotional activity following the publication of revised drinking guidelines

2. to use interrupted time series analysis of monthly survey data to assess whether or not trends in alcohol consumption behaviour, as a primary outcome, changed following publication and promotion of the revised drinking guidelines

3. to undertake subgroup analyses to examine whether or not there are variations in intervention effects across groups of the population defined by age, sex and socioeconomic status

4. to use difference-in-difference methods to examine whether or not direct and frequent exposure to promotion of drinking guidelines increases their effectiveness

5. to undertake pathway analyses to validate theorised capability, opportunity and motivation to change behaviour and behaviour itself

6. to assess cost-effectiveness of any identified effects on alcohol consumption using the Sheffield Alcohol Policy Model framework. 69

During the project, the objectives were amended as follows in consultation with the Project Steering Committee. First, as we identified no large-scale promotional activity relating to the guidelines beyond the announcement of the revisions, the first objective was abandoned. This was replaced with a new objective: to conduct a review of the scale and content of news media coverage relating to the guidelines. Second, preliminary analyses showed no changes in the primary outcome measure and, in the absence of any promotional activity, we decided not to conduct a cost-effectiveness analysis. Third, the lack of any substantial change in the outcome measures meant that we did not pursue the difference-in-difference analysis or pathway analyses.

At the same time, a decision was also taken to not conduct a cost-effectiveness analysis as preliminary analyses indicated that, in addition to there being no promotional activity that attracted a cost, there was also no change in the primary outcome (alcohol consumption). We judged, in consultation with the Project Steering Committee, that conducting the news media review was a better use of resources. The cost-effectiveness analysis is therefore not discussed further in this report.

Patient and public involvement

The research team conducted two patient and public involvement (PPI) sessions during the project to support the interpretation of the early and final results of the evaluation analyses. In the absence of a pre-existing and appropriate PPI panel in England, we worked with an established PPI panel based in Scotland. The panel comprised approximately 30 adult drinkers from the general population and living in or around Stirling, where the panel meets. The UK Centre for Tobacco and Alcohol Studies recruited the initial panel members in 2014 via social media, with additional members added over time to maintain the group’s size. The panel meets two or three times per year and provides members with a £25 voucher plus an offer to cover travel costs for each meeting attended.

One member of this PPI panel also joined the Project Advisory Group and attended the group’s meetings.

Structure of the report

The remainder of the report comprises a further four chapters. Chapter 2 describes the initial attempt to construct a timeline of promotional activity and the early findings prior to the abandonment of this exercise. Chapter 3 describes the review of print news media coverage of the guidelines. Chapter 4 describes the interrupted time series analyses. Chapter 5 brings together the main findings and discusses the strengths and limitations of the research and its implications for future research, policy and practice before presenting our conclusions.

Introduction

As stated in Chapter 1, the promotion of drinking guidelines is a complex intervention. The CMOs announced the UK Government’s revised guidelines in January 2016 and the Government’s Alcohol Strategy5 states that their purpose is to:

Support people to in make informed choices about healthier and responsible drinking.

Reproduced with permission. 5 © Crown copyright 2012. Contains public sector information licensed under the Open Government Licence v3.0

However, this is an incomplete account of the intervention and its aims. Subsequent promotional activity may be delivered by multiple stakeholders at different points in time, using different techniques and targeting outcomes that may be more or less proximal to alcohol consumption behaviours. These different promotional activities are likely to vary in their effectiveness, and changes in targeted outcomes arising from one activity may also be contingent on the timing and impact of other activities. As a result, identifying a single intervention time point or primary outcome measure for evaluation analyses is not straightforward.

Most previous studies evaluating the effectiveness of drinking guidelines as a public health intervention do not address this point. Many studies use simple before-and-after designs or analyses of annual time series and generic outcome measures, such as drinkers’ awareness of guidelines in general and their knowledge of specific guidelines. They also tend to evaluate the announcement of new or revised guidelines rather than the effects of specific promotional activities promoting those guidelines. As a result, there is little engagement with the actual aims, design or outcomes targeted by such activities. It is often unclear, therefore, whether promotion is ineffective, absent, poorly designed or targeted at outcomes that are not examined by evaluation studies. Exceptions to this include analyses of annual Danish promotional campaigns organised by the National Board of Health, which targeted different groups and used different messages in different years, and this is partly reflected in the interpretation of evaluation results. 37,38 Similarly, Dixon et al. ’s41 evaluation of a campaign in Australia that focused on drinking guidelines as well as alcohol and cancer in women surveyed samples of women on their knowledge of the alcohol–cancer link as well as more generic guideline-related knowledge and drinking behaviours before the campaign and at two time points afterwards.

For the purposes of this study, it is important that evaluation analyses start from a more detailed assessment of promotional activity. The use of a monthly time series within an interrupted time series design means that our results are likely to be sensitive to misspecification of the intervention point. For example, if the guidelines were announced in January 2016 but a large promotional campaign took place in September 2016, we would need to consider the appropriate analytical design to best estimate the effects of these two intervention points. It also clear that health promotion campaigns informed by theory are more effective than those with no clear theoretical basis and that behaviour change arises from multiple factors rather than a single outcome. 65 Therefore, it is important to assess the content and characteristics of promotional campaigns to identify their likely effectiveness, the outcomes in which we would expect to see changes and whether or not we would expect these changes, either alone or in combination with those arising from other promotional activities, to lead to changes in alcohol consumption behaviour.

We therefore aimed to construct a timeline of promotional activity relating to the drinking guidelines, detailing when activities occurred and what they involved. We also sought to identify the cost of activities to inform a planned cost-effectiveness analysis of promotional activity as a whole. For the purpose of this study, we defined promotional activity as any deliberate and planned effect to communicate the drinking guidelines with a view to having a national-level impact of a scale detectable within our analyses, whether directly (e.g. via a mass media campaign) or indirectly (e.g. via education of health professionals). We were interested in capturing promotion of any aspect of the guidelines described in the final documentation. 24 In addition to the guideline consumption threshold, this includes messages around drink-free days and the number of drinking days per week, the disputed cardioprotective benefits of moderate drinking, the cancer risks associated with low levels of alcohol consumption, risks for different adult age groups and drinking in pregnancy. The final example was not a focus of our study but was included in the data collection instruments for completeness.

Methods

Results

None of the five organisations reported undertaking any substantial promotional activity relating to the revised drinking guidelines between January 2016 and June 2017. Furthermore, none of the organisations reported any knowledge of substantial promotional activity undertaken by other organisations or any awareness of planned activities by their own or other organisations.

Some organisations did report smaller-scale activities. These included news media activities, such as issuing press releases or giving interviews. These activities usually related to the guidelines being announced in January 2016, their formal adoption in August 2016 or other occasional media requests.

Given the lack of substantial activity identified by our respondents or the research team after 19 months of our 22-month study and the lack of indications of any future activity, we decided that surveying the remaining organisations was unwarranted and we discontinued this work after discussions with our Project Advisory Group and the funder.

Discussion

The main finding of this work was that there was no substantial active promotion of the revised drinking guidelines after they were announced in January 2016. This does not mean that no organisation promoted the guidelines at all. It is clear that the guidelines are promoted intermittently on social media by a range of organisations, that organisations such as PHE and Drinkaware provide the guidelines and related information on their websites and in other materials, and that news media routinely mention the guidelines in day-to-day reporting of alcohol-related and health-related topics (see Chapter 3). However, it is unclear that these activities would substantially affect the design or findings of our evaluation analyses as they are unlikely to either reach large audiences or lead on their own to changes in behaviour or in key influences on behaviour change. 65 It is also unclear that they are of sufficient scale or sufficiently different in timing or nature to justify analysis of an additional intervention point beyond January 2016. Our monitoring of stakeholder websites and social media accounts suggests that most organisations changed their messaging on guidelines immediately after the January 2016 announcement and rolled out updated materials in the subsequent months. In the absence of large-scale promotional campaigns, it seems appropriate to regard this activity as part of a general accumulation of drinkers’ exposure to the revised drinking guidelines starting from January 2016 rather than seeking to identify additional intervention points.

Given the above, the remainder of this report proceeds on the basis that there was no significant activity promoting the revised guidelines after the announcement in January 2016. The implications of this for future research, policy and practice are discussed in the main discussion of the project findings in Chapter 5.

Introduction

News media can play an important role in communicating and shaping public perceptions of social issues, including those related to alcohol. However, the nature of this role is complicated and it is described in many different ways in the literature. For example, Casswell72 describes the media in general terms as ‘key players’ and argues that ‘much of the [public] debate is shaped by the portrayal of alcohol and alcohol policy in mass media’,72 a view broadly shared by several other authors. 73–75 Hansen and Gunter76 suggest that social issues are ‘elaborated, contested, redefined and eventually removed from the public agenda’76 within news media, implying that news media are engaged in multiple sequential or simultaneous processes. Törrönen77 views newspapers in more agentic terms, characterising their editorials as ‘indicators of the opinion climate . . . try[ing] consistently to exert influence . . . evaluate arguments and decisions . . . and make an assessment of them and of situations and attempting to influence policy and public opinion’,77 whereas Nicholls73 addresses only the first aspect of this quotation, noting that ‘news reporting plays a key role in articulating shared cultural values around alcohol’. Mercille78 argues more forcefully that news media deliberately, strategically and inevitably ‘. . . reflect the views of the political and economic establishment on public health measures’. Mercille78 argues this based on the links between media owners and government, their financial links to health-harming industries via advertising and the reliance on establishment contacts for sourcing of news stories and reactions. These different perspectives suggest that the role of news media in alcohol policy debate is complex and multifaceted. As such, it is difficult to capture or demonstrate fully within a single analysis,79 but remains an important area of study when assessing how public perceptions are formed on a particular topic.

Studies of news media coverage of alcohol use are a relatively recent phenomenon. In 2011, Nicholls73 noted that few such studies existed, particularly in the UK; however, a much larger body of literature emerged in subsequent years. This includes analyses of news media reports of policy debates around MUP in the UK,80–82 advertising restrictions and regulation of the night-time economy in Australia,74,83 strategies tackling college drinking in the USA84 and general alcohol policy in several countries. 77,78,85,86 It also includes analysis of news media coverage of contemporary social issues specific to alcohol or with which alcohol intersects, including alcohol’s harm to those other than the drinker,87 alcohol-related health risks,88,89 sex,75,90–93 sexual violence,94,95 parenting96 and personal freedoms. 97 Furthermore, a number of studies examine overall reporting of alcohol-related topics within news media. 73,98–101

A review of this literature in 2007 argued that analyses of news media coverage of alcohol, in addition to being few in number, had failed to draw on relevant theories and techniques developed within the communications and political science literature. 76 In particular, it noted the value of analyses of change over time, a lack of cross-referencing between studies, the mediating role of advocacy within news media or perspective taking by news organisations, the significance of message sources (e.g. individuals quoted within reports) and the medium of the coverage. More recent work has drawn on these arguments to develop rich understandings of how alcohol-related issues are reported in the media. For example, studies have demonstrated that public health actors have successfully positioned themselves as key sources of information within policy debates but provide conflicting findings as to whether or not those actors have also successfully framed debates in public health terms as population-level issues rather than the industry’s favoured individual-level and personal responsibility oriented perspective. 73,74,80 Many studies also identify key frames through which alcohol-related issues are discussed, many of which are replicated across policy debates. For example, men’s drinking is often found to be framed in relation to violence, whereas women’s drinking is linked to a wider set of concerns pertaining to transgression of traditional models of femininity. 73,75 Similarly, in debates about advertising, liberalisation of retail monopolies and MUP, opposition to alcohol policy is framed with reference to individual freedoms or the need to protect industry from an overbearing state. 74,77,80 Analyses of policy debates in Finland and Ireland also suggest that, as agentic actors in policy debate, news media are often unsupportive of public health arguments and alcohol-related problems; even when such problems are identified, the discussion and policies recommended tend to coincide with an underlying view that liberalised alcohol markets are desirable. 77,78

Despite this now-broad literature, only one briefly reported study to date has examined how news media report on drinking guidelines. 102 This is an important oversight as TV, radio, newspapers and magazines are the primary places where the public report seeing drinking guidelines when revisions are announced. 46 At such times, guidelines are discussed prominently by news outlets within ordinary news reports, opinion pieces and editorials. At other times, news media continue to expose the public to drinking guidelines by mentioning them within stories about alcohol-related health risks, drinking trends and lifestyle topics. Such repeated exposure aids awareness and recall of guidelines in the absence of active promotion via mass media campaigns. For example, evaluations of alcohol health warning labels in the USA found higher rates of awareness and recall of the health warnings among heavier drinkers and argued that this was because they see the label more often. 103 Debate within news media may also inform public opinion and responses to drinking guidelines, but, given the wide variety of roles that the media play in public debate, this is unlikely to be a straightforward or unidirectional process. 79 Moreover, differences in news media reports across publications and platforms may play an important role in determining who is exposed to guidelines and informing their attitudes towards them as engagement with news outlets and content is socially patterned. 76

Therefore, the aim of this review was to conduct an exploratory analysis to describe the scale and content of news media coverage of drinking guidelines in England during our study period of February 2014 to October 2017. As this was a retrospective study, broadcast media reports were difficult to access. Therefore, we focused on reviewing print news media, which are searchable via online databases. There were three stages of the review:

1. a systematic search to identify all news articles published in England during the study period that mentioned drinking guidelines or closely related terms

2. a quantitative analysis to develop a timeline of when and where the identified articles were published and to summarise the main content of articles

3. a qualitative analysis to explore the content of articles in more depth.

Methods

Results

Identified articles

Figure 2 shows the results of the search and selection process. The three Nexis searches yielded 10,983 articles in total. After excluding articles published in local publications or The Associated Press planners (i.e. collections of press releases that are disseminated to journalists and news publications, which are included in the Nexis database; n = 7070) and removing duplicates (n = 1702), we were left with 2211 articles in 41 publications. After concurrent headline and full-text screening of these articles, we excluded a further 1261 articles because they did not discuss the guidelines. The Google searches yielded 70 results on the BBC News website and six on the Sky News website. After headline and full-text screening of these articles, we excluded 29 articles (28 from the BBC) because they did not mention the guidelines. This left 41 further articles in addition to those identified via the Nexis search. After adding the website and Nexis searches together, our final data set comprised 997 eligible articles from 29 publications (Table 1).

FIGURE 2.

The PRISMA flow diagram showing the results of the search for news articles mentioning drinking guidelines.

TABLE 1 - Number of articles covering the UK alcohol guidelines in the total and stratified samples across publication and source type

Source type	Publication	Number of articles
		Total (n = 997)	Sample (n = 500)
Broadsheet/quality press	The Daily Telegraph (including telegraph.co.uk)	112	53
	The Independent (including the i, The Independent Magazine, independent.co.uk)	83	41
	The Guardian (including The Observer)	56	32
	The Times (including thetimes.co.uk, The Sunday Times)	88	44
Middle market	Daily Mail (including MailOnline, The Mail on Sunday)	215	99
	Daily Express (including Express Online, Sunday Express)	115	66
Red-top tabloid	Daily Mirror (including mirror.co.uk, The People)	62	29
	Daily Star (including Daily Star Online)	29	14
	The Sun (including thesun.co.uk, The Sun on Sunday)	96	52
Online only	BBC News	42	21
	BigHospitality.co.uk	1	0
	FoodNavigator.com	2	1
	Sky News	13	8
Other	Arts and Book Review	1	0
	Convenience Store	1	1
	Financial Wire	12	5
	Good Housekeeping	9	5
	Men’s Health	4	1
	Metro	11	5
	Morning Advertiser	8	3
	New Scientist	3	2
	Nursing Times	3	1
	Prima	5	2
	Runner’s World	1	0
	The Grocer	12	8
	The MCA Report (formerly The M&C Report)	2	1
	UK Government News	11	6

Quantitative analysis

Drinking guidelines were mentioned regularly in news articles across the study period. Figure 3 shows that the number of articles per week peaked when the revised guidelines were announced in January 2016. There were also many smaller peaks that occurred when guidelines were mentioned in articles related to general alcohol- or health-related topics, as shown in Figure 4, whereas guidelines were mentioned only infrequently in articles unrelated to alcohol or health.

FIGURE 3.

Number of news articles mentioning guidelines per week within the analysed sample (n = 500) between 1 February 2014 and 31 October 2017. The guidelines changed in week 2 of 2016.

FIGURE 4.

Number of news articles mentioning guidelines per week within the analysed sample (n = 500) between 1 February 2014 and 31 October 2017, split by primary topic in article. (a) Not related to drinking guidelines of alcohol or health; (b) related to drinking guidelines; (c) related to alcohol but not drinking guidelines; and (d) related to health but not alcohol.

The guidelines were mentioned for different reasons at different time points. Figure 5 shows that a large number of articles mentioned the guidelines in January 2016 to inform the public of the revisions and that there were also a large number of articles that discussed the merits of the guidelines at this time and in the subsequent 23 weeks. However, outside this period, the merits of the guidelines were rarely discussed, and guidelines were mentioned instead to promote health, to prevent ill health and, less often, to provide context to articles discussing alcohol consumption. Around the time that the revised guidelines were announced, there were also a number of articles that mentioned the guidelines for other reasons, including to discuss state intervention in health care and nanny state policies in general, and to criticise UK drinking culture.

FIGURE 5.

Number of news articles mentioning guidelines per week between 1 February 2014 and 31 October 2017, split by primary purpose of guidelines in article. (a) To discuss the merit of guidelines; (b) to inform the public of the change in guidelines; (c) to promote health; (d) to provide context for drinking habits; and (e) other.

Figure 6 shows that the tone of articles was typically neutral in the overwhelming majority of cases and that articles with a positive tone were rare. Articles with a negative tone were also rare but became more common for several weeks after the revisions were announced.

FIGURE 6.

Numbers of news articles mentioning guidelines per week between 1 February 2014 and 31 October 2017, split by article tone.

Table 2 shows the different aspects of the drinking guidelines that were mentioned in articles and how the aspects mentioned changed before and after the guidelines were revised. Most articles (95.4%) mentioned the guideline threshold, although sometimes this was provided for only one sex. Even after the same guidelines were set for men and women in January 2016, 94.0% of articles mentioned the guideline consumption level but only 88.6% mentioned that it applied to men. Prior to the revisions, the 1995 guidelines of 3–4 units per day for men and 2–3 units per day for women were mentioned in approximately half of the articles, but the pre-1995 guidelines of 21 units per week for men and 14 units per week for women were also mentioned in around one-third of the articles, with both guidelines sometimes mentioned for comparison (e.g. 3–4 units per day or 21 units per week), suggesting that the public was provided with inconsistent information during this time. After the revisions, only 9 out of 296 articles mentioned a guideline other than 14 units per week for men or women. Out of these nine articles, three mentioned the previous daily guidelines, whereas the other six mentioned the previous weekly guidelines.

TABLE 2 - Aspects of the guidelines that were mentioned before and after the guidelines change

Percentages are the proportion of articles that mention each aspect, and aspects are not mutually exclusive. Only aspects communicated in at least 5% of articles are mentioned.

Aspect of guidelines communicated	Before (n = 185)		After (n = 315)		Total (n = 500)
	Yes (%)	No (%)	Yes (%)	No (%)	Yes (%)	No (%)
Recommended consumption levels (at least one of the following)	97.8	2.2	94.0	6.0	95.4	4.6
3–4 units per day for men	56.8	43.2	0.6	99.4	21.4	78.6
14 units per week for men	6.5	93.5	88.6	11.4	58.2	41.8
21 units per week for men	35.7	64.3	1.9	38.1	14.4	85.6
2–3 units per day for women	59.5	40.5	1.0	99.0	22.6	77.4
14 units per week for women	36.8	63.2	88.9	11.1	69.6	30.4
Spread weekly limit over multiple days	1.1	98.9	13.7	86.3	9.0	91.0
No alcohol on at least 2 days per week	9.2	90.8	10.2	89.8	9.8	90.2
No safe level of drinking
Communicated as part of guidelines	2.2	97.8	23.2	76.8	15.4	84.6
Communicated separate from guidelines	4.9	95.1	2.5	97.5	3.4	96.6

Other aspects of the guidelines were mentioned much less frequently, and the announcement of the revisions did not always lead to increased mentions. The advice to spread the weekly guideline of 14 units across at least 3 days was mentioned in 13.7% of articles after the announcement, compared with 1.1% before. However, the advice to have at least 2 drink-free days per week was mentioned in 10.2% of articles after the revisions and 9.2% before, perhaps because this advice was part of PHE’s Change for Life mass media campaign prior to the guideline review.

Information that was not part of the guidelines was also sometimes presented as though it was. One particular example of this arose because the revised guidelines were announced in conjunction with a report on the relationship between alcohol consumption and cancer, which argued that there was no safe level of drinking with regard to cancer risk. 105 Although the UK CMOs’ report and the report of the GDG discussed evidence on alcohol and cancer and noted the risk at low levels of consumption, the ‘no safe level’ message is not part of the drinking guidelines. 22,24 Despite this, 23.2% of articles published after the revisions stated or implied that this message was part of the guidelines, for example by mentioning it shortly before or after the consumption guidelines without clarifying that it is not part of the guidelines. This compares with 2.5% of articles that mentioned the ‘no safe level’ message as separate to the guidelines (e.g. by stating this explicitly, or by attributing the ‘no safe level’ message to a source other than the guidelines).

Relationship between timing, tone and other article characteristics

We examined the relationship between the tone of articles, whether they were published before or after the guidelines announcement, and their other characteristics. As only 11 out of 500 articles analysed were positive in tone, the analysis focuses on a comparison of articles with a neutral or negative tone.

Table 3 shows that articles were significantly more likely to have a negative tone and less likely to have a neutral tone after the revised guidelines were announced (χ² = 20.3; p < 0.001). Across the whole study period, articles that had drinking guidelines as the primary topic were more likely to have a negative tone than those related to alcohol in general, those related to health in general or those that were unrelated to guidelines, alcohol or health (χ² = 179.2; p < 0.001). Guidelines were also more likely to be discussed negatively if the primary purpose of mentioning the guidelines was to discuss their merits (χ² = 282.8; p < 0.001). There were no significant differences in the likelihood of the guidelines being discussed negatively across source types (χ² = 4.76; p = 0.31) or their political leanings (χ² = 0.66; p = 0.72). Articles were, however, more likely to be negative in tone if they communicated the ‘no safe level’ message as part of the guidelines rather than identifying it as a separate message or not mentioning it at all (χ² = 42.32; p < 0.001).

TABLE 3 - Context in which guidelines were communicated between 1 February 2014 and 31 October 2017

Context in which guidelines were communicated	Positive (%)			Negative (%)			Neutral (%)			Total (%)
	Before (n = 1)	After (n = 10)	Total (n = 11)	Before (n = 10)	After (n = 62)	Total (n = 72)	Before (n = 174)	After (n = 243)	Total (n = 417)	Before (n = 185)	After (n = 315)	Total (n = 500)
Source type
Broadsheet/quality press	0.0	40.0	36.4	50.0	32.3	34.7	36.8	31.7	33.8	37.3	32.1	34.0
Middle market	100.0	30.0	36.4	40.0	29.0	30.1	27.6	37.4	33.3	28.6	35.6	33.0
Red-top tabloid	0.0	0.0	0.0	0.0	21.0	18.1	23.0	17.3	19.7	21.6	17.5	19.0
Online only	0.0	30.0	27.3	10.0	1.6	2.8	2.9	8.2	6.0	3.2	7.6	6.0
Other	0.0	0.0	0.0	0.0	16.1	13.9	9.8	5.3	7.2	9.2	7.3	8.0
Context
Related to drinking guidelines	0.0	60.0	54.5	40.0	80.6	75.0	6.3	10.7	8.9	8.1	26.0	19.4
Related to alcohol (but not drinking guidelines)	0.0	40.0	36.4	40.0	11.3	15.3	69.5	67.9	68.6	67.6	55.9	60.2
Related to health (but not alcohol)	100.0	0.0	9.1	10.0	4.8	5.6	19.5	17.7	18.5	19.5	14.6	16.4
Not related to guidelines, alcohol or health	0.0	0.0	0.0	10.0	3.2	4.2	4.6	3.7	4.1	4.9	3.5	4.0
Role of guidelines within article
To discuss merit of guidelines	0.0	30.0	27.3	50.0	82.3	77.8	0.0	6.2	3.6	2.7	21.9	14.8
To inform public of change in guidelines (but not discuss their appropriateness)	0.0	20.0	18.2	20.0	3.2	5.6	6.9	6.6	6.7	7.6	6.3	6.8
To promote health	100.0	30.0	36.4	20.0	3.2	5.6	59.8	56.8	58.0	57.8	45.4	50.0
To provide context for drinking	0.0	0.0	0.0	0.0	1.6	1.4	28.2	22.2	24.7	26.5	17.5	20.8
Other	0.0	20.0	18.2	10.0	9.7	9.7	5.2	8.2	7.0	5.4	8.9	7.6

Discussion

Mentions of drinking guidelines in UK news media articles during our study period took one of three general forms. First, there were incidental mentions within articles about broader alcohol- or health-related topics. These mentions accounted for 76.6% of articles, occurred on a continuous basis over time, and were generally informational in purpose and neutral in tone. Second, guidelines were mentioned in articles reporting and explaining the revisions to the guidelines. These mentions accounted for 19.4% of articles, occurred immediately after the announcement in January 2016, and were also generally informational in purpose and neutral in tone. Third, articles mentioned the guidelines when discussing their merits. These mentions accounts for 4.0% of articles, occurred primarily in the 23 weeks after the revisions were announced, and were discursive in purpose and often negative in tone.

As mentions of guidelines in most articles were either incidental or factual, these provided little scope for thematic analysis. Therefore, our themes emerged primarily from the third type of article and generally relate to criticism of the guidelines. These criticisms focus on the guidelines’ scientific basis and the integrity of the process by which they were developed, the broader purpose of guidelines with regard to a potentially malign public health agenda and the perceived inconsistency of public health advice around alcohol and other behaviours.

The main strengths of this analysis are that it covers periods when revised guidelines were discussed in the news media and status quo periods, whereas many previous articles focused only on a specific policy debate. It also combined quantitative and qualitative analyses to provide a detailed account of how guidelines are communicated to the public through news media. The analysis also has important limitations. First, we did not examine broadcast or social media as it was not feasible to do so within this retrospective analysis. This is common in many studies of news coverage of alcohol policy questions. 74,76–78,80,81 In part, this is because newspaper coverage is of importance in its own right as a distinct form of news reporting, but it is also the case that newspapers in the UK broadly reflect and shape the news that is reported by broadcast outlets. Therefore, we would not expect our findings to vary substantially were we to conduct a more resource-intensive review of broadcast news stories. Furthermore, UK broadcast news is dominated by the BBC, and we captured an understanding of its reporting via inclusion of articles on the BBC News website, as well as the website of another major broadcaster, Sky News. A previous study136 examined initial reactions to the guidelines by tracking a single hashtag on Twitter (www.twitter.com; Twitter, Inc., San Francisco, CA, USA) for 1 week immediately after they were announced and found a roughly equal split between tweets that were supportive and unsupportive of the guidelines. Second, we searched only for articles that mentioned the guidelines and, therefore, we cannot assess how much alcohol-related news the public is exposed to that does not make mention of the guidelines, nor can we compare the coverage of drinking guidelines with other alcohol-related advice that may be reported (e.g. drinking clear spirits to limit calorie intake from alcohol). Third, we do not know the social patterning of the audience for different articles. Identifying such information would be challenging as media outlets promote certain articles more than others to attract advertising revenue or because of newsworthiness. Thus, although neutral articles were more common, it may be that these articles attracted a different audience or were less read overall than more negative articles. The analyses in Chapter 4 do, however, provide some insights into whether or not the overall body of news reports the impact on key outcomes relating to drinking guidelines. Fourth, our search focused only on the main guideline message to not consume more than 14 units per week. Other elements of the guidelines, such as those relating to drink-free days or spreading consumption across multiple days, were not covered and we may, therefore, have missed some aspects of the news media’s discussion of the guidelines.

The findings suggest that the revised guidelines, including some of the secondary advice (e.g. on drink-free days), were initially communicated to the public via large-scale news media coverage, which was mostly factual and accurate. This communication was then reinforced on a continuous basis by incidental mention of the guidelines within alcohol- and health-related articles. If this communication was effective, we would expect changes in proximal outcome measures (e.g. awareness and knowledge of the guidelines) to accumulate sharply following the initial announcement and then continue to accumulate more steadily in the subsequent time period as the public is exposed repeatedly to the guidelines via incidental mentions in news articles. However, there are three caveats to this. First, health messages require reinforcement and effects may decline among drinkers who are exposed to the initial announcement but not to subsequent mentions of guidelines in lower-profile news stories that are less likely to be read by certain sociodemographic groups. 76 Second, diffusion theory suggests that repetition of a message does not lead to universal penetration. 52 Instead, some people may never be exposed to the message and others may become insensitive to it because of overexposure. According to theories of psychological reactance,58 some drinkers may go further and actively reject the message or choose to increase their drinking if they perceive it as an unjustified threat to their freedom, as is described in some of the nanny state-oriented criticisms. Third, the negative commentary on the revised guidelines and the substantive criticisms that underpin it may also lead some people to reject or misunderstand the guidelines even if they are exposed to them. In combination, these caveats may mean that the initial sharp accumulation in effects on proximal outcomes from the announcement of revisions dissipates among drinkers who are not further exposed to the guidelines. Effects on these outcomes may then begin to accumulate steadily as some drinkers are repeatedly exposed to the guidelines through incidental mentions within articles, but this accumulation may plateau as the maximum level of diffusion is reached. Effects on distal outcomes, such as alcohol consumption behaviour, may be unlikely to arise from the news media reports identified as only the initial reporting contained sufficiently detailed explanation of the guidelines to constitute a call to action. Incidental mentions were more likely to state only the guideline consumption level and, therefore, were unlikely to prompt behaviour change. In the absence of any large-scale promotional campaign beyond the initial announcement, we would not expect this trajectory to be interrupted by any sharp changes in outcome measures at further time points after the initial announcement of revisions.

The findings also point to important implications for practice. First, the lack of articles presenting a positive, as opposed to simply factual, view of the drinking guidelines suggests a one-sided public debate. The lack of positive accounts of the guidelines from regular newspaper columnists is perhaps unsurprising as the majority of UK newspapers are regarded as on the political right and are often hostile to public health interventions. However, it is commonplace for public health actors to engage in public debate within newspapers via opinion pieces and letters to editors,73 so it is unclear why this did not happen after the revised drinking guidelines were announced. Analyses of previous policy debates in the UK and internationally suggest that, although public health actors have successfully embedded themselves with news media debate as valued information sources, their contributions are often less co-ordinated or strategic than those of commercial actors. 74,80 The dominance of negatively framed discussion of the guidelines after the initial announcement may be a further example of this strategic deficit.

Second, the criticism of drinking guidelines can be looked at from two perspectives. On the one hand, it features strategies commonly used by commercial actors within policy debate. These features include cherry-picking of evidence (e.g. citing evidence of cardiovascular benefits without noting that these were included in the epidemiological model underpinning the guidelines),137 invoking complexity as a barrier to action (e.g. highlighting that risk varies between individuals)138 and, less widely discussed in prior literature, questioning the integrity and motives of public health actors. On the other hand, public health actors should be conscious of their active role in facilitating these criticisms. For example, the decision to announce the revised guidelines alongside a report on cancer played a critical role in the guidelines being conflated with the ‘no safe level’ message, one of the key points of contention for critics. The CMO reinforced this by telling a parliamentary committee that people should follow my lead. Every time I drink, I choose between my glass of wine or an increased risk of breast cancer (paraphrased). 139 In a radio interview, she also described the extensive if disputed scientific evidence showing cardiovascular benefits of moderate drinking as ‘an old wives’ tale’ (The Guardian, 8 January 2016; reproduced with permission; Courtesy of Guardian News & Media Ltd). 140 Similarly, several members of the GDG are prominent advocates for public health interventions to tackle alcohol-related harm and can, therefore, legitimately be regarded as both political and scientific actors. 21 Furthermore, the public is presented with conflicting alcohol-related health advice on a regular basis, usually by scientific and public health institutions keen to promote their work. 141 Previous studies highlight the resulting difficulties faced and strategies adopted by the public in using such conflicting information to construct their understandings of health-related risks. 127,131,142 These examples suggest that criticism of drinking guidelines is in part a strategic response by commercial actors to a threat to their interests, but is also a reflection of the way that public health actors, and particularly public health scientists, engage with news media and the wider policy debate.

Third, the day-to-day reporting of alcohol-related topics is a continuous process irrespective of major policy announcements and presents an opportunity for exposing the public to alcohol-related health information. If this is regarded as important by public health actors, then they should seek to embed such information within press releases or in supporting or reaction quotations. Although the effectiveness of techniques used by different actors is not tested directly, previous research suggests that embedding multiple arguments within single quotations may be effective as well as providing a greater focus on population-level perspectives, the harms of alcohol to drinkers and others and the likely benefits of individual policies and multicomponent strategies. 74,80 In undertaking these actions, public health actors should, however, be mindful of the problems outlined above with regard to public health actors’ role in constructing criticism of the guidelines.

Introduction

There is little robust evidence available on if and how the promotion of drinking guidelines affects alcohol-related outcomes. 13,28 One reason for this is that health authorities often do not actively promote drinking guidelines after publishing them. Our analysis of promotional activity (see Chapter 2) and news media coverage (see Chapter 3) suggests that this was the case for the 2016 revisions to the UK guidelines. We identified no large-scale promotional activity by the UK Government, PHE or any third-sector organisation. We also identified no large peaks in news media coverage after the initial announcement of the revised guidelines in January 2016.

The lack of promotional activity has three implications for our evaluation. First, it indicates that we should focus on January 2016 as our intervention point. Although the government did not formally adopt the revised guidelines until it had completed a public consultation in August 2016, there was no substantial change in news reporting at this time, and this leaves the January announcement as the only large-scale promotional event. Second, the lack of any large-scale promotional activity after January 2016 means that we cannot undertake our planned analysis of the effect of promotional campaigns that occurred at different points in time and we treat January 2016 as the sole intervention point. Third, our preregistered protocol and earlier chapters of this report state that our aim is to evaluate the impact of promoting revised drinking guidelines on alcohol consumption behaviour. However, in the absence of significant promotional activity, it is more accurate to describe the analyses below as an evaluation of the impact of announcing such guidelines, and we will use this language henceforth.

We planned to use a monthly repeat cross-sectional survey of adults in England to conduct a series of analyses to evaluate the impact of announcing the guidelines in three stages. First, we describe trends in respondents’ awareness and knowledge of drinking guidelines, where they report being exposed to the guidelines and influences on change in drinking behaviours. As part of this, we test whether or not the average level of measures of these outcomes differed between the pre- and post-intervention time point (i.e. January 2016) and across key population subgroups defined by sociodemographics and alcohol consumption level. Second, the primary analysis uses segmented regression to test whether or not alcohol consumption behaviour changed following the announcement of the revised drinking guidelines. Third, we assess the robustness of our results and competing interpretations via a series of secondary analyses involving alternative outcome measures and model specifications.

Methods

Results

Descriptive analyses: awareness, knowledge and exposure

Table 5 summarises the characteristics of the sample and provides pre- and post-intervention data on the primary and secondary outcome measures, as well as measures of statistical differences in measures between these time points. Figure 7 shows that the proportion of respondents who were drinkers varied across the study period, between 72.8% in April 2015 and 62.7% in November 2016. The proportion of respondents who were drinkers was higher in the pre-intervention period than in the post-intervention period (70.4% vs. 65.7%, χ² = 171.5; p < 0.001). During the shorter period from November 2015, when the ATS included questions related to drinking guidelines, the proportion of respondents who were drinkers peaked at 70.3% in November 2015.

TABLE 5 - Number of respondents and differences between the pre- and post-intervention periods for all measures

Unweighted numbers and weighted column percentages.

Measure collected from March 2014. All other measures collected from November 2015.

Pre- and post-intervention figures are means and test statistic is the t-statistic.

Moderate drinkers are those consuming < 14 units per week. Increasing-risk drinkers are men consuming between 14 and 50 units per week and women consuming between 14 and 35 units per week. High-risk drinkers are men consuming > 50 units per week and women consuming > 35 units per week.

Bold text indicates that differences were significant at the 95% confidence level.

Adapted with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Measure	Total (n)a	Pre interventiona	Post interventiona	Chi-squared test	p-value
Total
n (March 2014 to October 2017)	74,388	36,905	37,483
n (November 2015 to October 2017)	40,832	3349	37,483
Sex
Men	74,371	49.0	49.0	0.0	0.99
Women		51.0	51.0
Age (years)
16–34	74,388	31.1	31.1	0.3	0.73
35–64		48.1	47.9
≥ 65		20.8	21.1
Social grade
AB	74,388	27.0	27.1	0.1	0.93
C1C2		49.4	49.3
DE		43.6	23.7
Drinker (March 2014 – October 2017)	48,875	70.4	65.7	163.7	0.00
AUDIT-C score (mean)b	48,696	4.1 c	4.4 c	–10.2 c	0.00
Litres of ethanol released for sale per adult		0.81c	0.79c	0.64c	0.26
Assaults		2261.7c	2197.6c	1.0c	0.16
Alcohol poisonings		190.9c	186.2c	0.7c	0.25
Drinker (November 2015 to October 2017)	26,335	68.1	65.7	6.5	0.01
Units per week from graduated frequency	22,404	9.7	9.9	–0.3	0.61
Moderated	21,829	78.3	77.3	0.9	0.41
Increasing riskd		18.7	20.0
High riskd		3.0	2.7
Aware of guidelines	26,306	86.0	87.0	1.6	0.20
Identifies guidelines as 14 units per week	22,466	21.4	23.9	16.6	0.00
Last month exposure to guideline via:	19,548
Product label		20.1	22.1	2.6	0.11
TV or radio		34.2	39.1	11.8	0.00
Newspapers or magazines		15.8	17.6	2.5	0.11
Websites or social media		6.0	7.5	3.7	0.05
Shops or supermarkets		8.3	8.3	0.0	0.98
Pubs, bars or nightclubs		13.0	11.1	4.5	0.03
Place of work or study		6.5	7.4	1.1	0.30
Health professional		8.9	8.0	1.5	0.22
Health service		11.1	11.7	0.5	0.47
Friends, family or colleagues		6.4	7.5	2.0	0.16
Other		1.4	0.6	9.4	0.00
None of these places		29.3	23.5	20.9	0.00
COM-B measures
Can drink only up to 2 units before harm	24,245	34.2	35.6	1.26	0.26
Easy to drink moderately	26,203	82.5	82.0	0.24	0.63
Tracks units	26,274	26.7	25.9	0.46	0.50
Lifestyle makes it easy	26,143	76.9	79.4	6.5	0.01
Know where to get advice	26,233	78.6	82.3	16.0	0.00
Trying not to drink more than is good for me	26,063	48.7	51.2	3.8	0.05
Intends to drink moderately	26,249	77.0	77.5	0.2	0.65
Wants to avoid drinking excessively	26,247	76.0	76.4	0.2	0.69
Trying to avoid drinking excessively	26,224	39.3	39.6	0.1	0.77
Concerned by drinking too much	26,256	24.4	23.8	0.3	0.56

FIGURE 7.

The proportion of respondents who are (a) drinkers; (b) drinkers who are aware of the guidelines; and (c) drinkers who can correctly identify the guideline. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Figure 7 also shows that awareness of guidelines among drinkers was high across the shorter study period, at an average of 86.9%, and did not change substantively after the revised guidelines were announced. Furthermore, Figure 7 shows that knowledge of guidelines did change to a modest degree during the study period. Among drinkers who were aware of guidelines, in December 2015, 20.6% said that the guideline was 14 units per week (or 2 units per day), and this rose to 29.0% in January 2016, when the guidelines were announced. However, this increase in knowledge dissipated over subsequent months, and only 24.1% of drinkers identified the guideline correctly in October 2017.

Overall, the proportion of drinkers who correctly identified the guideline was slightly lower in the pre-intervention than the post-intervention period, suggesting a small improvement in knowledge (20.6% vs. 23.7%, χ² = 4.2; p = 0.04). The analyses show a similar pattern of results if those who said that the guideline was < 14 units are combined with those who gave the correct answer.

The guideline for men changed to a greater degree than the guideline for women, so we compared the change in knowledge between the two sexes (Figure 8). Among drinkers, in December 2015, 9.1% of men who were aware of the guidelines said that the guideline was 14 units per week, and this rose to 23.8% in January 2016 before falling to 20.1% in October 2017. Among women, the change was smaller: the proportion of drinkers who stated that the guideline was 14 units per week was 33.2% in December 2015, rising to 34.6% in January 2016 and falling to 28.6% in October 2017. The proportion of male drinkers identifying the guideline correctly was higher in the post-intervention period than in the pre-intervention period (17.6% vs. 9.1%, χ² = 42.9; p < 0.001), but this was not the case for women (30.1% vs. 31.7%, χ² = 0.9; p = 0.35).

FIGURE 8.

The proportion of drinkers correctly identifying the guidelines, by sex.

Exposure questions were asked only to the 68.8% of drinkers who were able to name a figure for the guideline consumption level irrespective of whether or not it was correct. There were only three locations in which more than 20% of this subsample reported exposure to the guideline in the last month (Figure 9). These locations were on the TV or radio, on product labels, and in newspapers or magazines. TV and radio was the most commonly cited source of exposure to the guideline, with 32.5% reporting last-month exposure in December 2015, rising to 64.8% following the guideline announcement in January 2016, which attracted prominent news coverage. Exposure via TV and radio then fell gradually to vary between 28.8% and 35.6% across 2017. Exposure via newspapers and magazines showed a similar peak and then decline but varied only between 12.6% and 24.3% across the study period. As would be expected, exposure via product labels did not peak when the revised guidelines were announced as product labels were not updated immediately and exposure via labels varied between 14.3% and 27.7% across the study period. Overall, slightly more of this subsample of drinkers reported exposure to their reported guideline figure post intervention than pre intervention (76.5% vs. 70.8%, χ² = 20.9; p < 0.001). However, a significant increase in exposure was found only in the case of TV and radio (39.1% vs. 34.2%). The findings on exposure did not differ substantively when limiting the subsample to those respondents who correctly identified the guideline rather than those identifying any guideline. Figure 10 shows that, across the study period, only between 20% and 30% of the subsample were not exposed to guidelines in any location during the last month.

FIGURE 9.

The proportion of drinkers giving a figure for the guidelines who reported exposure to that figure in different locations during the last month. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

FIGURE 10.

The proportion of drinkers giving a figure for the guidelines who reported no exposure to that figure in any location during the last month.

There were differences between sociodemographic and consumption groups in likelihood of being a drinker as well as awareness, knowledge and exposure to the guideline (Figure 11; see Appendix 3 for full regression results). Respondents who were men, older or of a higher social grade were more likely to be drinkers. In general, awareness, knowledge and exposure were higher among those in higher social grades and those drinking at increasing- or high-risk levels. For example, awareness and knowledge varied particularly by social grade. Awareness fell from 92.4% among those in grade AB to 78.5% among those in grade DE, and 27.3% of those in grade AB identified the guideline correctly, compared with 18.1% of those in grade DE. Exposure showed more variation in social patterning depending on the medium, although rates of exposure were generally low in all cases and the most likely means of exposure for all groups were via product labels, TV or radio, and newspapers or magazines. Overall, those aged 16–24 years and heavier drinkers were more likely to report exposure in at least one medium. Older drinkers were more likely to report exposure through TV, radio, newspapers and magazines but younger drinkers were more likely to report exposure via product labels, licensed venues or talking to friends, family and colleagues. Those in lower social grades were less likely to report exposure via newspapers and magazines. Notably, heavier drinkers were more likely to report exposure by talking to health professionals, or from posters or booklets at health services, as well as by talking to friends, family or colleagues.

FIGURE 11.

The proportion of respondents who are drinkers, aware of guidelines, know the correct guideline or have been exposed to the guidelines in different locations, by sociodemographic and consumption group. (a) Drinker; (b) aware of guidelines; (c) know correct guideline; (d) exposed via product labels; (e) exposed via TV or radio; (f) exposed via newspapers or magazines; (g) exposed via websites or social media; (h) exposed via shops or supermarkets; (i) exposed via pubs, bars or nightclubs; (j) exposed via place of work or study; (k) exposed via health professionals; (l) exposed via health services; (m) exposed via friends, etc.; (n) exposed via other means; and (o) exposed in no places. Light-blue bars indicate a significant difference compared with the left-most category within characteristic in multivariate logistic regressions adjusting for all listed characteristics.

Descriptive analyses: capability, opportunity and motivation

Figures 12–15 show trends in the dichotomised COM-B measures over the study period. Although some of these measures did change significantly after the announcement of revised guidelines, these changes were small and there is little evidence of any substantial change over time.

FIGURE 12.

The proportion of respondents answering positively to capability-related outcome measures over time. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

FIGURE 13.

The proportion of respondents answering positively to opportunity-related outcome measures over time. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

FIGURE 14.

The proportion of respondents answering positively to automatic motivation-related outcome measures over time. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

FIGURE 15.

The proportion of respondents answering positively to reflective motivation-related outcome measures over time. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Capability

Across the whole study period, 35.5% of drinkers thought that they could regularly drink no more than 2 units per day before doing significant harm to their health. This did not change significantly after the announcement of revised guidelines (34.25% pre change vs. 35.6% post change, χ² = 1.26; p = 0.26). Unlike the knowledge question above, there was also no change among male drinkers (24.8% pre change vs. 26.9% post change, χ² = 1.67; p = 0.20). Most drinkers indicated that they would find it easy to drink ≤ 3 units per day, with 82.0% saying this across the study period, although only 26.0% indicated that they often tracked their units. In both cases, there was no significant change between the pre- and post-intervention periods.

Opportunity

The proportion of drinkers who indicated that their lifestyle made it easy to drink ≤ 3 units per day was 79.2% across the study period and increased slightly after the announcement of revised guidelines (76.9% pre change vs. 79.4% post change, χ² = 6.5; p = 0.01). The proportion who knew where to get advice to cut down drinking was also high across the study period, at 82.0%, and, again, increased slightly post intervention (78.6% pre change vs. 82.3% post change, χ² = 16.0; p < 0.001).

Motivation

Only a minority of drinkers (23.9%) were concerned about drinking more alcohol than was good for them, but 51.0% of drinkers wanted to avoid consuming this amount, although 76.4% wanted to keep their drinking within safe limits. Only the proportion wanting to avoid drinking more than was good for them increased post intervention, and the increase was small (48.7% vs. 51.2%, χ² = 3.9; p = 0.05). A large minority, comprising 39.6% of drinkers, were trying not to drink more than was good for them across the study period, and 77.4% indicated that they intended to drink within safe limits. Neither of these two measures changed post intervention.

Figure 16 shows that there are several significant differences in the COM-B measures between sociodemographic and, particularly, drinking groups during the post-intervention period (see Appendix 3 for full regression results).

FIGURE 16.

The proportion of respondents giving positive responses to COM-B questions by sociodemographic and consumption. (a) Knowledge: 1–2 units is most can consume on a day before it does significant harm to health; (b) perceived capability: easy to drink ≤ 3 units per day; (c) skills: often keeps track of units; (d) social opportunity: lifestyle makes it easy to drink ≤ 3 units per day; (e) social opportunity: knows where to get advice on cutting down on drinking; (f) automatic motivation: concerned about drinking more alcohol than is good for you; (g) automatic motivation: wants to avoid drinking more than is good for you; (h) reflective motivation: trying to avoid drinking more than is good for you; (i) automatic motivation: wants to keep drinking within safe limits; and (j) reflective motivation: intends to drink within safe limits. Light-blue bars indicate a significant difference compared with the left-most category within characteristic in multivariate logistic regressions adjusting for all listed characteristics.

Although the descriptive analyses above show that men’s knowledge of the correct guideline increased over the study period, women were significantly more likely than men to say that ≤ 2 units per day is the most they can consume regularly before it harms their health. Differences on other measures, although statistically significant in most cases, were smaller. The exceptions are that women were more likely than men to both want and intend to keep their drinking within safe limits.

With regard to age, the largest differences were that those aged ≥ 65 years were more likely to say that they would find it easy to drink ≤ 3 units per day and that their lifestyle made this easy. This age group was also more likely to want to, and intend to, keep their drinking within safe limits. Differences between those aged 16–34 years and those aged 35–64 years were generally small, although often statistically significant. The largest difference was that those aged 16–34 years were less likely to say that their lifestyle made it easy to drink < 3 units per day.

Differences by social grade were of similar size to those by age. Those in social grade AB were more likely than those in lower social grades, and particularly those in grade DE, to say that they would find it easy to drink ≤ 3 units per day, that their lifestyle made it easy to do so and that they knew where to get advice on cutting down drinking. They were also more likely to say that they wanted to avoid drinking more than is good for them and both wanted and intended to stay within safe limits.

The differences between drinker groups were much larger than the sociodemographic differences. Although the descriptive analyses above show that increasing- and high-risk drinkers are more likely to know and be exposed to the guidelines, the COM-B measures suggest that they are much less likely to say that ≤ 2 units per day is the most they can consume regularly before it harms their health. Increasing-risk and especially high-risk drinkers are also much less likely to say that they are able to drink ≤ 3 units per day or that their lifestyle would make this easy. Similarly, increasing-risk and especially high-risk drinkers were much less likely to say that they wanted to or were trying to avoid drinking more than is good for them or that they wanted to or intended to keep their drinking within safe limits.

Primary analyses (planned): alcohol consumption

The primary evaluation analyses showed that there was no significant change in the level of AUDIT-C scores in the unadjusted model after the announcement of the revised guideline, but there was a significant change in the AUDIT-C score trend (Figure 17 and Table 6). In December 2015, the modelled average AUDIT-C score was 2.719, and it was decreasing by 0.003 each month. After the January 2016 announcement, AUDIT-C scores increased immediately but non-significantly to 2.720 [β = 0.001, 95% confidence interval (CI) –0.079 to 0.099] and the trend changed significantly such that scores subsequently increased by 0.005 each month (β = 0.008, 95% CI 0.001 to 0.015). This is equivalent to an additional 0.5% of the population increasing their AUDIT-C score by 1 point each month by, for example, drinking four or more times per week rather than two or three times per week. The results did not change after adjusting for average temperature, but the change in trend was no longer significant when controlling for alcohol prices or in the fully adjusted model.

FIGURE 17.

Primary analysis showing AUDIT-C scores before and after the announcement of revised drinking guidelines in January 2016. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

TABLE 6 - Results of the segmented regression analysis assessing the impact of the new drinking guidelines on AUDIT-C scores

Bold text indicates that differences were significant at the 95% confidence level.

Adapted with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Coefficient	Unadjusted			Adjusted for temperature			Adjusted for off-trade and on-trade sales of beer and wine			Adjusted for temperature and on-trade and off-trade sales of beer and wine
	β	95% CI	p-value	β	95% CI	p-value	β	95% CI	p-value	β	95% CI	p-value
Intercept	2.785	2.723 to 2.847	< 0.001	2.740	2.554 to 2.926	< 0.001	5.055	–5.893 to 16.003	0.365	3.484	–7.775 to 14.743	0.544
Pre-intervention trend	–0.003	–0.008 to 0.002	0.306	–0.002	–0.007 to 0.003	0.325	< 0.001	–0.010 to 0.010	0.975	< 0.001	–0.010 to 0.010	0.977
Step-level change	0.001	–0.079 to 0.099	0.824	0.011	–0.078 to 0.100	0.805	0.027	–0.081 to 0.135	0.619	0.029	–0.079 to 0.137	0.596
Change in trend	0.008	0.001 to 0.015	0.015	0.008	0.001 to 0.015	0.020	0.007	–0.008 to 0.022	0.343	0.008	–0.007 to 0.023	0.314
Temperature				0.004	–0.012 to 0.020	0.615				0.010	–0.007 to 0.027	0.241
Wine on							–0.002	–0.003 to –0.001	0.002	–0.002	–0.003 to –0.001	0.002
Wine off							0.023	–0.003 to 0.049	0.083	0.026	< 0.001 to 0.052	0.053
Beer on							–0.020	–0.105 to 0.065	0.649	–0.011	–0.097 to 0.075	0.810
Beer off							–0.020	–0.049 to 0.009	0.178	–0.019	–0.048 to 0.010	0.185

Secondary analyses (planned)

Tables 7 and 8 show the results of analyses using alternative graduated frequency and taxation measures, respectively (also shown in Figures 18 and 19). There was a non-significant increase in alcohol consumption measured by graduated frequency of 0.1 units after the announcement of revised guidelines (β = 0.105, 95% CI –0.652 to 0.862). Similarly, there was an immediate non-significant increase in per capita litres of alcohol released for sale of 0.006 l (β = 0.006, 95% CI –0.025 to 0.036), and the trend in this measure changed non-significantly from declining by 0.001 l per capita per month to remaining stable (β = 0.001, 95% CI –0.001 to 0.004).

TABLE 7 - Results of the segmented regression analysis assessing the impact of the new drinking guidelines on graduated frequency

Coefficient	Unadjusted			Adjusted for temperature			Adjusted for off-trade and on-trade sales of beer and wine			Adjusted for temperature and on-trade and off-trade sales of beer and wine
	β	95% CI	p-value	β	95% CI	p-value	β	95% CI	p-value	β	95% CI	p-value
Intercept	9.861	9.134 to 10.588	< 0.001	9.737	8.882 to 10.592	< 0.001	–0.567	–58.095 to 56.961	0.985	7.858	–62.354 to 78.070	0.826
Step-level change	0.105	–0.652 to 0.862	0.785	0.089	–0.671 to 0.849	0.818	0.047	–0.715 to 0.809	0.905	–0.007	–0.799 to 0.785	0.987
Temperature				0.013	–0.034 to 0.060	0.588				–0.002	–0.068 to 0.064	0.955
Wine on							0.008	–0.001 to 0.017	0.088	0.008	–0.001 to 0.017	0.088
Wine off							–0.241	–0.483 to 0.001	0.051	–0.225	–0.478 to 0.028	0.081
Beer on							0.175	–0.304 to 0.654	0.474	0.114	–0.455 to 0.683	0.696
Beer off							0.074	–0.150 to 0.298	0.520	0.048	–0.193 to 0.289	0.699

TABLE 8 - Results of the segmented regression analysis assessing the impact of the new drinking guidelines on alcohol released for sale per capita

Coefficient	Unadjusted			Adjusted for temperature			Adjusted for off-trade and on-trade sales of beer and wine			Adjusted for temperature and on-trade and off-trade sales of beer and wine
	β	95% CI	p-value	β	95% CI	p-value	β	95% CI	p-value	β	95% CI	p-value
Intercept	0.797	0.776 to 0.818	< 0.001	0.77	0.679 to 0.861	< 0.001	7.696	–1.32 to 16.713	0.105	7.364	–2.242 to 16.97	0.144
Pre-intervention trend	–0.001	–0.002 to 0.001	0.405	–0.001	–0.002 to 0.001	0.451	–0.007	–0.015 to 0.001	0.116	–0.006	–0.015 to 0.003	0.174
Step-level change	0.006	–0.025 to 0.036	0.715	0.005	–0.026 to 0.037	0.747	–0.060	–0.144 to 0.024	0.174	–0.056	–0.144 to 0.032	0.223
Change in trend	0.001	–0.001 to 0.004	0.248	0.001	–0.001 to 0.004	0.294	0.003	–0.007 to 0.013	0.559	0.003	–0.008 to 0.014	0.577
Temperature				0.002	–0.005 to 0.010	0.548				0.006	–0.008 to 0.021	0.394
Wine on							0.001	< 0.001 to 0.002	0.093	0.001	< 0.001 to 0.002	0.071
Wine off							–0.025	–0.054 to 0.004	0.102	–0.020	–0.051 to 0.011	0.209
Beer on							–0.032	–0.101 to 0.038	0.379	–0.032	–0.104 to 0.041	0.399
Beer off							–0.014	–0.037 to 0.010	0.264	7.364	–2.242 to 16.97	0.144

FIGURE 18.

Mean weekly consumption in units measured by graduated frequency before and after the announcement of revised drinking guidelines in January 2016 (data only available from November 2015).

FIGURE 19.

Litres of ethanol released for sale per capita adult before and after the announcement of revised drinking guidelines in January 2016. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Tables 9 and 10 show the results for measures of alcohol-related hospital admissions for assaults and alcohol poisonings, respectively (also shown in Figures 20 and 21). There was a significant step change and change in trend for incidence of assault hospitalisations after the announcement of the revised guidelines. Hospitalisations for assault decreased by 7.3% immediately after the announcement [incidence rate ratio (IRR) 0.927, 95% CI 0.875 to 0.983], but this was partly counteracted by a change from a flat trend pre intervention to a 0.5% increase in the incidence of assault hospitalisations each month (IRR 1.005, 95% CI 1.001 to 1.009). The step change but not the change in trend remained in the adjusted models. There was also a significant step change of –15.4% in the incidence of alcohol poisoning hospitalisations (IRR 0.846, 95% CI 0.735 to 0.975), whereas the trend changed non-significantly from increasing by 0.9% each month to increasing by 0.5% each month (IRR 0.996, 95% CI 0.984 to 1.007). The step change was no longer significant after adjusting for the temperature trend.

TABLE 9 - Results of segmented regression analysis assessing the impact of the new drinking guidelines on incidence of assaults with adjustment for first-order autoregressive autocorrelation

The high estimates values for the intercept indicate an overparameterised model and results should be treated with caution.

Bold text indicates that differences were significant at the 95% confidence level.

Adapted with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Coefficient	Unadjusted			Adjusted for temperature			Adjusted for off-trade and on-trade sales of beer and wine			Adjusted for temperature and on-trade and off-trade sales of beer and wine
	IRR	95% CI	p-value	IRR	95% CI	p-value	IRR	95% CIa	p-value	IRRa	95% CIa	p-value
Intercept	2249	2160 to 2342	< 0.001	1786	1592 to 2004	< 0.001	4,885,231	4395 to 5,430,641,194	< 0.001	236,874	739 to 75,971,915	< 0.001
Pre-intervention trend	1.000	0.997 to 1.003	0.934	1.001	0.998 to 1.003	0.757	1.002	0.996 to 1.008	0.546	1.001	0.996 to 1.006	0.807
Step-level change	0.927	0.875 to 0.983	0.015	0.927	0.878 to 0.978	0.009	0.941	0.885 to 1.000	0.061	0.925	0.875 to 0.977	0.009
Change in trend	1.005	1.001 to 1.009	0.030	1.004	1.000 to 1.008	0.080	0.997	0.989 to 1.005	0.440	0.999	0.992 to 1.007	0.867
Temperature				1.021	1.011 to 1.031	< 0.001				1.015	1.012 to 1.019	< 0.001
Wine on							1.001	1.000 to 1.002	0.045	1.001	1.000 to 1.002	0.012
Wine off							0.985	0.967 to 1.003	0.103	0.984	0.971 to 0.998	0.031
Beer on							0.948	0.898 to 1.001	0.062	0.969	0.927 to 1.013	0.172
Beer off							1.005	0.987 to 1.024	0.594	1.006	0.990 to 1.021	0.471

TABLE 10 - Results of segmented regression analysis assessing the impact of the new drinking guidelines on incidence of alcohol poisonings with adjustment for first-order autoregressive autocorrelation

Bold text indicates that differences were significant at the 95% confidence level.

Adapted with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Coefficient	Unadjusted			Adjusted for temperature			Adjusted for off-trade and on-trade sales of beer and wine			Adjusted for temperature and on-trade and off-trade sales of beer and wine
	IRR	95% CI	p-value	IRR	95% CI	p-value	IRR	95% CI	p-value	IRR	95% CI	p-value
Intercept	173	156 to 192	< 0.001	149	129 to 172	< 0.001	86,236	0 to 41,597,607,191	0.098	6267	0 to 5,691,240,593	0.220
Pre-intervention trend	1.009	1.001 to 1.017	0.040	1.008	1.000 to 1.016	0.061	1.009	0.996 to 1.022	0.189	1.008	0.995 to 1.022	0.221
Step-level change	0.846	0.735 to 0.975	0.026	0.874	0.760 to 1.004	0.064	0.862	0.749 to 0.993	0.047	0.868	0.751 to 1.004	0.064
Change in trend	0.996	0.984 to 1.007	0.443	0.995	0.983 to 1.006	0.383	0.992	0.973 to 1.011	0.413	0.993	0.974 to 1.013	0.500
Temperature				1.015	1.006 to 1.023	0.002				1.010	0.998 to 1.023	0.123
Wine on							1.002	1.000 to 1.004	0.022	1.002	1.000 to 1.004	0.040
Wine off							1.000	0.972 to 1.029	0.994	1.001	0.973 to 1.029	0.975
Beer on							0.958	0.865 to 1.060	0.409	0.974	0.876 to 1.083	0.628
Beer off							0.987	0.954 to 1.022	0.467	0.991	0.957 to 1.027	0.629

FIGURE 20.

Number of hospitalisations for assaults before and after the announcement of revised drinking guidelines in January 2016.

FIGURE 21.

Number of hospitalisations for alcohol poisoning before and after the announcement of revised drinking guidelines in January 2016.

We used an iterative procedure to identify any statistically significant break points in the AUDIT-C score trend. Table 11 and Figure 22 show that the procedure identified a break point in June 2015, 7 months before the announcement of the revised guidelines. In this model, AUDIT-C scores declined by 0.005 points per month up to June 2015 and began to increase significantly, by 0.011 points per month, thereafter (β = 0.011, 95% CI 0.003 to 0.019).

TABLE 11 - Results of the iterative segmented regression analysis assessing the impact of the new drinking guidelines on AUDIT-C scores at alternative break points

Coefficient	β	95% CI	p-value
Intercept	2.806	2.739 to 2.873	< 0.001
Trend up to June 2015	–0.005	–0.011 to 0.001	0.059
Trend after June 2015	0.011	0.003 to 0.019	0.006

FIGURE 22.

Model-defined break point in trend in mean AUDIT-C score across the study period. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

We also tested for the presence of pulse effects at 2 and 3 months after the announcement of the revised guidelines. Table 12 shows that we found evidence in both cases of a significant pulse effect, such that AUDIT-C scores were significantly lower, by 0.184 points for 2 months after the intervention (β = –0.184, 95% CI –0.292 to 0.076) and by 0.126 points for 3 months after the intervention (β = –0.126, 95% CI –0.218 to 0.034). These effects are equivalent to 1.8% of the population reducing their AUDIT-C score by 1 point for 2 months or 1.3% of the population reducing their AUDIT-C score by 1 point for 3 months. We also ran unplanned sensitivity analyses to assess when the identified pulse effects dissipated. The results in Table 13 suggest that the pulse effects lasted for a 4-month period after the announcement of the revised guidelines. Figure 23 shows the pulse effect visually.

TABLE 12 - Results of the pulse regression analysis assessing the impact of the new drinking guidelines on AUDIT-C scores

Coefficient	β	95% CI	p-value
Intercept	2.740	2.697 to 2.783	< 0.001
Underlying trend	0.002	< 0.001 to 0.004	0.009
Pulse effect 2 months	–0.184	–0.292 to –0.076	0.001
Pulse effect 3 months	–0.126	–0.218 to –0.034	0.007

TABLE 13 - Results of the duration of pulse regression analysis assessing the impact of the new drinking guidelines on AUDIT-C scores

Bold text indicates that differences were significant at the 95% confidence level.

Adapted with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Coefficient	β	95% CI	p-value
Intercept	2.740	2.697 to 2.783	< 0.001
Underlying trend	0.002	< 0.001 to 0.004	0.009
Pulse effect 2 months	–0.184	–0.292 to –0.076	0.001
Pulse effect 3 months	–0.126	–0.218 to –0.034	0.007
Pulse effect 4 months	–0.087	–0.167 to –0.007	0.032
Pulse effect 5 months	–0.042	–0.114 to 0.030	0.246
Pulse effect 6 months	–0.035	–0.101 to 0.031	0.290

FIGURE 23.

Short-term or pulse effect on AUDIT-C following the announcement of revised drinking guidelines in January 2016. Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Reproduced with permission from Holmes et al. 143 This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

The AUDIT-C scores showed greater than anticipated variation across the post-intervention period, so we ran unplanned sensitivity analyses to assess whether or not the results changed when using an extended post-intervention period of 48 months (i.e. until February 2018). Table 14 shows that we found similar results to the primary analysis. The pre-intervention trend of a reduction in AUDIT-C scores of 0.002 units each month changed significantly to an increase in AUDIT-C scores of 0.010 units each month (β = 0.010, 95% CI 0.004 to 0.016). Furthermore, in planned sensitivity analyses, we tested whether quadratic or cubic segmented regressions models provided a better fit.

TABLE 14 - Results of the segmented regression analysis assessing the impact of the new drinking guidelines on AUDIT-C scores with an extended post-intervention period

Coefficient	β	95% CI	p-value
Intercept	2.785	2.722 to 2.848	< 0.001
Pre-intervention trend	–0.002	–0.007 to 0.003	0.342
Step-level change	–0.015	–0.099 to 0.069	0.731
Change in trend	0.001	0.004 to 0.016	0.001

Discussion

The analyses above suggest that announcing revised drinking guidelines did not lead to a step-change decline or a reduction in the longer-term trend in self-reported alcohol consumption behaviour. Further analyses inconsistently indicated small and short-term reductions in various secondary outcomes relating to alcohol consumption. There were small and selective impacts on drinkers’ awareness and knowledge of the guidelines and key influences on behaviour change. We summarise the results below and then discuss the key strengths and limitations of the analysis. Chapter 5 presents a broader summary of the project findings, drawing on previous chapters, as well as a discussion of the implications for policy, practice and future research of the main evaluation analyses presented here.

Summary of findings

The overall findings of this project are that the announcement of the revised UK lower-risk drinking guidelines (1) was not subject to large-scale promotion by any organisation following the initial media launch, (2) was not associated with substantial and sustained changes in most influences on behaviour change and (3) was not associated with clearly detectable changes in drinking behaviour. In this chapter, we expand on and discuss these three findings in turn.

The drinking guidelines were not subject to large-scale promotion by any organisation following the initial announcement

Announcing revised guidelines did not lead to substantial and sustained changes in most influences on behaviour change

Analyses of trends in key influences on behaviour change, specifically drinkers’ awareness, knowledge and exposure to drinking guidelines and their capability, opportunity and motivation to change their behaviour, suggest that none of these outcomes changed substantially or in a sustained way following the announcement of revised guidelines. This is perhaps unsurprising given that large-scale promotional activity was limited to a short period of earned news media coverage rather than a well-designed campaign aiming to change drinkers’ behaviour or affect their preparation to change their behaviour. Nonetheless, the lack of substantial changes in drinkers’ awareness and, especially, knowledge of the guidelines merits discussion.

Over 85% of drinkers were aware of guidelines before the announcement and, although this did not increase after the announcement, this figure is a high baseline. However, fewer than 25% of drinkers were able to correctly identify the guideline, and fewer than half gave a figure below 14 units per week or 2 units per day. These findings are broadly in line with a separate nationally representative UK survey conducted 1 month after the revised guidelines were announced, which found that just 8% of drinkers knew the revised guidelines. 182 In the present study, the proportion of men who knew the guideline did increase significantly after the revision was announced but remained below 25% at the end of the study period. The equivalent figure for women was only 29%, despite the guideline for women changing to a lesser degree, suggesting a general lack of knowledge of drinking guidelines among both sexes. It is unlikely that the government’s aim for drinkers to use the guidelines to make responsible and healthy choices about alcohol consumption will be achieved with such low rates of knowledge about what the guideline is.

Responses to other questions assessing influences on behaviour change varied. Among the COM-B measures of drinkers’ capability, opportunity and motivations to change behaviour, most drinkers felt that it was easy to drink at moderate levels (i.e. ≤ 3 units per day), felt that their lifestyle made it easy to do so, wanted to do so, intended to do so and said that they knew where to get help if they wanted to cut down. However, drinkers gave less positive responses on two capability measures. First, only around one-quarter of drinkers kept track of their units. Although unit counting may be unnecessary for those already drinking moderately, the proportion doing so was even lower among drinkers exceeding the guidelines. Second, only around one-third of drinkers said that 1–2 units is the most they can consume on 1 day before it does significant harm. This aligns with the findings above on drinkers’ low knowledge of the guideline, but it also suggests that drinkers may require convincing that the guideline is set at an appropriate level. Drinkers also had mixed motivations for behaviour change. Most drinkers wanted to and intended to stay within safe limits but only half of drinkers were trying not to drink more than is good for them and, as with tracking units, this was more common among those already drinking within the guidelines than among those exceeding them. Only one-quarter of drinkers were concerned by drinking more alcohol than is good for them, and < 40% were trying to avoid drinking excessively. The contradiction between these responses and drinkers’ wanting to and intending to stay within safe limits suggests that, although, in a general sense, drinkers are motivated to stay within the guidelines, they are not strongly motivated by concerns about alcohol-related health risks or necessarily willing to forgo heavier drinking. This appears to reflect a concern expressed repeatedly in prior literature regarding a conflict between the consumption threshold set by guidelines and the desire of drinkers to become intoxicated. 23,42,142 This, in turn, reflects a broader concern that drinking guidelines, like other forms of health promotion, engage insufficiently with positive motivations for unhealthy behaviour, such as pleasure, transgression and the enhancement of traits such as self-confidence, and we return to this point below. 183,184

In addition to examining time trends in behavioural influences, we also examined the social patterning of these influences and identified important inequalities. In particular, drinkers in social grade DE were the least likely to be aware of guidelines or to know what the guideline is. They also found it more difficult to drink moderately, were less motivated to do so and were less sure where to get advice on cutting down. Taken together, these findings suggest that drinkers of lower socioeconomic status are less able and less likely than their counterparts of higher socioeconomic status to use the revised drinking guidelines to make responsible and healthy choices about alcohol consumption. As drinkers of lower socioeconomic status also suffer greater harm from alcohol despite drinking less, on average,161,185 it is likely that these differences will contribute to the perpetuation and, possibly, exacerbation of alcohol-related health inequalities.

Furthermore, drinkers exceeding the guidelines also differed significantly from those not doing so on the measures of behavioural influences. Heavier drinkers were more likely to be aware of and exposed to the guidelines, potentially reflecting an accumulation effect from seeing guidelines on product labels and more general engagement with or sensitivity to alcohol-related materials. However, these heavier drinkers were no more likely than moderate drinkers to know what the guideline was. It is difficult to explain this anomaly. It may be a form of psychological reactance, in which individuals reject or disengage from information that threatens their freedom to act. 58 Support for this interpretation is provided by the finding that heavier drinkers were also substantially less likely to say that 2 units per day is the most they can drink without harming their health. As with the low level of agreement on this measure across the population, it suggests that heavier drinkers in particular require convincing that the drinking guidelines are set at an appropriate level. Heavier drinkers were also less likely to respond positively to other questions about their capability to change behaviour. They were less likely than those drinking within the guidelines to find it easy to drink at moderate levels and were less likely to track the units they consume. They also responded less positively on most measures of their motivations.

In general, these findings suggest that those at elevated risk from their drinking are among the least well prepared to use the drinking guidelines to make informed or healthier choices about their alcohol consumption.

Announcing revised guidelines did not lead to clearly detectable changes in drinking behaviour

The main evaluation analyses present a somewhat mixed picture of the impact of announcing the revised guidelines on alcohol consumption behaviours, with different findings for different outcome measures and analytical models. Analyses of our primary outcome measure suggest that the revised guidelines were associated with an increasing trend towards more problematic alcohol consumption, but this trend appears to have begun 6 months prior to the revised guidelines being announced and it may also have been only a short-term trend. The findings were also specific to our primary outcome measure of AUDIT-C scores and were not seen in our secondary measures of drinking behaviours derived from graduated frequency questions and alcohol taxation data. There was evidence of a step change to a lower level of hospital admissions for alcohol poisoning and assaults following the guideline announcement, but these changes were too large to be attributed plausibly to the guidelines given the lack of evidence of change in any other outcome measures. We judge that these changes in hospitalisations are more likely to result from other uncontrolled confounding factors or from artefacts of the research methods.

These findings are in line with previous studies examining the effects of alcohol-related health information on alcohol consumption, including previous announcements of new or revised drinking guidelines, evaluation or testing of health warning labels and evaluation of mass media campaigns. 28,41,49,63 They are also in line with our logic model, shown in Appendix 1, which suggests that changes in alcohol consumption behaviours are contingent on promotional activity, which was limited, and changes in influences on behaviour change, which did not take place.

Strengths and limitations

Patient and public involvement

We conducted two PPI sessions throughout the course of the project to support the interpretation of the early and final results of the evaluation analyses. The sessions took place in 2016 and 2018. In the absence of a pre-existing and appropriate PPI panel in England, we worked with an established PPI panel comprising approximately 30 adults within the general population of Scotland who regularly consume alcohol. The lead researcher gave a presentation on drinking guidelines, the present project and the preliminary findings at that point in time. He then led a discussion of panel members’ views on drinking guidelines, their exposure to the revised guidelines and their perspective on the results.

The main outcome of these sessions was to confirm our interpretation that the guidelines had had little impact on drinkers in the general population. The comments of panel members generally matched our own interpretations of the extent of promotional activity, the likelihood of the guidelines affecting behaviour and the reasons underlying this. For example, some panel members said that they had noticed the news reports announcing the new guidelines but had seen little other information and had largely forgotten the content of those news reports. Panel members were apathetic towards the guidelines for a variety of reasons, including their impenetrability, scepticism about the ‘nanny state’ and their lack of usefulness for those motivated to drink to intoxication. These views aligned well with the findings of our survey data and the themes we discuss above when interpreting our results. Panel members also suggested that the messaging they had seen, particularly in news media, was generally negative and focused on the risks of drinking. They were keen to see more positive messages that engaged with drinkers’ motivations to drink or the benefits of not drinking. They also wanted to see messaging that set out the positive actions that drinkers could take to gain these benefits.

The PPI sessions were useful to the extent that they confirmed our interpretations of the data, but they did not substantially influence the methods or write-up of the study as no new practicable suggestions or ideas for research activities or interpreting the data and results emerged. Panel members were engaged and insightful in their comments and the format of the sessions worked well.

One panel member also attended Project Advisory Group meetings and made useful suggestions regarding sources of news media data.

Implications for policy

The findings of this project suggest that the public is not in a position to use the revised drinking guidelines to make the responsible and healthy choices about their drinking envisaged by the 2012 Alcohol Strategy. 5 This finding echoes those from previous studies relating to the 1995 UK guidelines40,186 and to guidelines internationally. 33,36 Although governments aspire to use drinking guidelines and unit or standard drink information to help drinkers to consume alcohol in moderation, they have not been successful in equipping drinkers with the knowledge and skills to achieve this or in persuading drinkers that the guideline is appropriate or useful. Below, we discuss three key areas of concern: problems with units, problems with the guidelines in general and problems with the specific revised UK guideline.

First, the available evidence suggests that drinkers in the UK have only a partial understanding of units and few use units to monitor their consumption. 23,34,40,47,146 Drinkers typically overestimate the size of 1 unit, suggest that it is difficult to remember the unit content of drinks when this information is not provided (e.g. in on-trade settings) and argue that it is difficult to use the information in real-life contexts. 8,34,42,195 This is problematic as units are the metric by which guidelines communicate the recommended maximum level of consumption to the public. However, the problem is slightly more complicated than it appears at first glance. The UK Opinions Survey found that 90% of adults had heard of units in 2009 and that, although only one-quarter of respondents knew that 1 unit of wine was less than a 125-ml glass, two-thirds knew that 1 unit of beer was half a pint and 1 unit of a spirit was a single measure, and half said that 1 unit of wine was a ‘small glass’. 40 This suggests that the basic knowledge of units is less of a problem than the skills necessary to apply that knowledge in practice, for example to estimate the number of units in a particular drink, to adapt estimates to beers, wine or spirits of different alcoholic strengths or to calculate the total number of units consumed across a number of differently sized drinks drank during a night out. Equipping people with these skills is likely to be important if drinkers are to effectively use units as part of an effort to moderate their alcohol consumption.

Second, the evidence from the present project and previous research is also clear that drinkers have only a limited understanding of drinking guidelines, which, coupled with an incomplete understanding of units, is likely to hinder their ability to respond to guidelines as envisaged by the government’s alcohol strategy. The present project additionally demonstrates that drinkers’ capabilities, opportunities and motivations to use the guidelines to change behaviour are inconsistent. Key areas for attention include drinkers’ knowledge of the guidelines, their beliefs regarding the riskiness of exceeding the recommended maximum consumption level, and aspects of their motivations relating to deliberate efforts to drink at moderate levels and avoid excessive drinking. The need for greater knowledge and preparedness to use the information to change behaviour is also not evenly distributed across the population. In particular, those drinkers at greatest risk from their alcohol consumption, namely heavier drinkers and those in lower socioeconomic groups, differ in important ways from those at lower risk. Our findings suggest that enabling heavier drinkers to change their alcohol consumption behaviours will require increases in their motivations to do so and also changes in their beliefs about the riskiness of higher levels of alcohol consumption. Those in lower socioeconomic groups require greater awareness and knowledge of guidelines if they are to use them effectively and, if they are to change their drinking behaviour, efforts may also be required to influence their motivations to do so.

Third, evidence that drinkers believe that they can drink > 2 units per day without being at risk of serious harm suggests that the UK guideline of 14 units per week does not align well with drinkers’ perceptions of risky levels of alcohol consumption. This may reflect that the guideline is set to recommend a low-risk level of consumption rather than to identify a high-risk level. However, it may also reflect drinkers’ perceptions of the risks associated with lower levels of alcohol consumption. For example, several studies have demonstrated low awareness of the relationship between alcohol consumption and cancer risk,196–198 which is an important driver of risks at low consumption levels. There is also widespread awareness among drinkers of evidence that low levels of alcohol consumption may provide some protection against cardiovascular disease,199,200 despite the most recent evidence suggesting that such protective effects may be smaller and less certain than previously thought. 201–203 Furthermore, qualitative evidence suggests that drinkers’ judgements regarding the risks of exceeding the guidelines are informed less by their understanding of epidemiological data and more by a wider set of considerations that jointly amount to a lay epidemiology of alcohol. 23,131,204 These considerations do include formal scientific evidence, as reported via news media, but also include drinkers’ understandings of alcohol-related risks as developed through their own experience of alcohol consumption and alcohol-related problems and the consequences of alcohol consumption they observe and discuss within their families, social networks and communities. Drinkers’ judgements also draw on embodied understandings of risk, regarding whether they feel they are intoxicated or unwell or have reached their personal consumption limit. 23,42,142,187,205 In combination, this mixture of inconsistent awareness of epidemiological evidence and risk perceptions informed by a wide set of personalised factors presents two options for increasing drinkers’ use of guidelines. First, health authorities may seek to increase public understanding of alcohol-related health risk and increase their motivations to avoid these risks. One approach to this may be to promote the basis for the guidelines, that drinking 14 units per week entails an approximately 1% risk of dying due to alcohol (although note that we have argued elsewhere that this basis for the guidelines is problematic188). The 1% figure could be set in context by comparing it to the risk associated with other everyday activities. However, the effectiveness of this approach may be limited to the extent that drinkers find that information does not accord with their own experience and lay epidemiological perspectives and their positive motivations for drinking. 184 The second option, therefore, is to develop further evidence on how drinking guidelines can be designed and prompted. This evidence should address how drinkers use, or might use, guidelines in practice. 188

Implications for future research

We identify three priority areas for future research, namely identification of effective alcohol-related health promotion messages, how to tailor those messages to be effective for particularly at-risk groups and how to set drinking guidelines to be effective as a behaviour change intervention. We discuss each of these in turn.

There is an extensive literature describing effective approaches to changing behaviour, but there is a lack of evidence to inform the specifics of what alcohol-related health promotion messages should say if they are to be effective. For example, a review by Hassan and Shiu64 found that the literature on the effectiveness of warning labels for alcoholic products was ‘weak and fragmented’,64 with little consistency in research approach and few clear recommendations available on key questions such as the use of imagery, recommendations and specific versus general messages. 64 The small literature evaluating the effectiveness of promoting drinking guidelines similarly provides little guidance on which messages are more or less effective, often because evaluation studies do not focus on particular campaigns or do not describe those campaigns in detail. Addressing these problems through a more structured programme of research that promotes comparability across studies and focuses on key unanswered questions is an important priority going forwards.

A corollary to this is the need to identify messages that are effective for particular population groups. This project identifies heavier drinkers and those of lower socioeconomic status as groups of particular concern as both groups are at elevated risk from their alcohol consumption and less prepared to change their behaviour. The existing literature offers little guidance on how to tailor alcohol-related health promotion messages to these groups. Studies evaluating the mandatory US alcohol warning labels do describe diffusion processes by which heavier drinkers, who are exposed to the labels through increased contact with alcoholic products, are more aware of and able to recall the warning labels. 52 This literature also demonstrates that there are limits to diffusion of awareness and knowledge via the simple accumulation of exposure, as trends in these outcomes plateaued 3–4 years after the labels were implemented. 52 Study authors often discuss the possibility that this reflects psychological reactance or the effects of repeated exposure that no longer attracts attention,58,64,136 but there is little direct evidence to support these arguments. We are unaware of any evidence discussing the effectiveness of alcohol-related health promotion messages targeting those of lower socioeconomic status. As above, developing an evidence base in this area is an important priority.

Finally, greater attention is needed to the design of drinking guidelines themselves. As we have discussed elsewhere, guidelines are often designed with reference, or at least explicit reference, only to formal epidemiological evidence. 188 This is particularly true of the guideline consumption threshold. Little consideration or evidence is brought to bear on questions regarding whether or not guidelines and the consumption threshold are appropriate to achieve the desired effect. This omission is particularly concerning where guidelines are conceived as a behaviour change intervention, as in the UK,5 but remains important where the purpose of guidelines is to communicate scientific evidence to the public, as reducing an extensive and nuanced evidence base to a single number is inevitably a simplification of that evidence and different approaches may communicate the full implications of the evidence in better or worse ways. The EU-funded Reducing Alcohol Related Harm project made best practice recommendations for the development of drinking guidelines in 2017,8 but these appear to be based primarily on epidemiological considerations and do not engage with considerations drawn from the wider scientific literature, such as how drinkers actually use guidelines in practice, how guidelines intersect with positive motivations for drinking (e.g. pleasure) and how the effectiveness of drinking guidelines is shaped by their acceptability and credibility within public debate.

Conclusions

The announcement of revised UK low-risk drinking guidelines was not subject to large-scale promotion by any organisation following the initial media launch, was not associated with substantial and sustained changes in most influences on behaviour change and was not associated with clearly detectable changes in drinking behaviour. Therefore, we conclude that the limited promotional activity does not appear to have led to the guidelines being used to inform responsible and healthier choices among drinkers, as intended by the UK Government.

Contributions of authors

John Holmes (https://orcid.org/0000-0001-9283-2151) (Reader in Alcohol Policy, Public Health) led the project and drafted the report.

Emma Beard (https://orcid.org/0000-0001-8586-1261) (Research Fellow, Health Psychology and Statistics) conducted the evaluation analyses and contributed to the drafting of the report.

Jamie Brown (https://orcid.org/0000-0002-2797-5428) (Principal Research Fellow, Health Psychology) contributed to project and analysis design, evaluation analysis and drafting of the report.

Alan Brennan (https://orcid.org/0000-0002-1025-312X) (Professor of Health Economics and Decision Modelling, Health Economics) contributed to project and analysis design and drafting of the report.

Inge Kersbergen (https://orcid.org/0000-0002-8799-8963) (Research Associate, Public Health) led the review of news coverage of the drinking guidelines and contributed to drafting of the report.

Petra S Meier (https://orcid.org/0000-0001-5354-1933) (Professor of Public Health, Public Health) contributed to project and analysis design and drafting of the report.

Susan Michie (https://orcid.org/0000-0003-0063-6378) (Professor of Health Psychology, Health Psychology) contributed to project and analysis design and drafting of the report.

Abigail K Stevely (https://orcid.org/0000-0002-5637-5245) (PhD Student, Public Health) analysed influences on behaviour change and contributed to drafting of the report.

Penny Buykx (https://orcid.org/0000-0003-4788-4002) (Associate Professor, Behavioural Science) contributed to project design, led the review of promotional activity and contributed to drafting of the report.

Publication

Holmes J, Beard E, Brown J, Brennan A, Meier PS, Michie S, et al. Effects on alcohol consumption of announcing and implementing revised UK low-risk drinking guidelines: findings from an interrupted time series analysis [published online ahead of print July 19 2020]. J Epidemiol Community Health 2020.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to available anonymised data may be granted following review.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health and Social Care.