Whilst technically true – as any level of alcohol does increase the risk of some types of cancers for example – the message has not been universally supported within the public health field.
For one, ‘no safe level’ may be seen as in some way contradictory to lower risk drinking guidelines which reflect that drinking at low levels doesn’t usually cause serious harm. Relatedly, some of the pushback
against the ‘no safe level’ message has centred on the principle that many things we do carry risk, whether that be eating processed food or driving a car, but we are not expected to avoid them altogether just because they involve some potential level of future harm. Other concerns have been raised about how the message is sometimes invoked in ways that appear to position all alcohol use as simply ‘wrong’.
Despite this, the ‘no safe level’ message is now widespread, in part due to the World Health Organisation’s promotion of it, but has also been adopted by other people or organisations speaking out about alcohol harm. Indeed, highlighting there is ‘no safe level’ might intuitively feel like a good way to counter persuasive alcohol industry messaging and widespread normalisation and acceptability of alcohol use that can drive harm.
However, one question that has been raised is what is the actual purpose of a ‘no safe level’ message? Is it intended to promote behaviour change amongst alcohol users (if so, which ones?), or perhaps increase support for tighter policies to reduce alcohol harm? Or something else? Just because a message might intuitively feel powerful and address an important knowledge gap, doesn’t mean it is necessarily effective.
Our study sought to test how the ‘no safe level’ message was interpreted by those drinking above low risk levels (of no more than 14 units a week) as those most likely to experience alcohol-related harm. We recruited this sample because, despite them being the most likely to benefit from reducing or stopping their drinking, these groups might also be the hardest to engage with via such messaging. In a previous study, we found that the more people drank, the more they rejected an alcohol health risk infographic, and the more they underestimated how susceptible they’d be to alcohol-related harms.
These psychological processes are sometimes called ‘defensive processing’ and tend to operate subconsciously, often in response to emotions such as fear. We therefore also included a gain-framed ‘benefits of cutting down’ message that did not contain negative risk framed information such as cancer risk, which would be expected not to produce fear and defensive processing. We also included a neutral control condition, and measured whether these messages affected problem recognition (how much people thought their drinking could be considered a problem) and risk recognition (how much risk they thought there was of drinking above guidelines (including to things like certain cancers)).
In our study, the ‘no safe level’ message included a script which repeated this headline message, whilst also giving further information about the basis of the message including cancer risks. Our results were interesting, but also quite complex. For recognising one’s own drinking as a problem, the ‘no safe level’ message did appear to increase problem recognition compared to the control group, but this effect was small, and also was not significant when compared against the benefits of ‘cutting down’ message.
This could be read as there being some potential for a ‘no safe level’ to increase problem recognition amongst those who the message has the most relevance to, but certainly not convincing evidence. Further, neither message increased personal risk recognition of drinking above the guidelines, including cancer and other health-related risks.
Perhaps the most interesting findings related to the defensive processing measures. The ‘no safe level’ substantially increased feelings of fear, and relatedly, led to people derogating the message’s credibility (i.e., they perceived the message as more untrustworthy, exaggerated etc.). These effects were far larger for the ‘no safe level’ than the ‘benefits of cutting down’ message, and provide strong evidence that such messages can evoke defensive processing responses that may undermine the effectiveness of such message for those they are most relevant to.
Another important finding was that both messages decreased self-efficacy. Self-efficacy - the extent to which a person believes that they can change a behaviour – is an especially important component of alcohol-related behaviour change. Indeed, past research shows how low self-efficacy often leads to defensive processing – if we think we can’t change a behaviour, there is no use in worrying about it. This may be why graphic health warnings might not be as effective in prompting behaviour change as we might intuitively think.
On the broader level, these findings highlight the importance of recognising psychological factors such as self-efficacy and defensive processing when using alcohol-related messages. Indeed, previous research has argued that building self-efficacy should be prioritised alongside any efforts to increase understanding of alcohol-related risks. If we can boost people’s belief that change is possible, we can increase their openness to considering the benefits of doing so.
Our results don’t necessarily mean a ‘no safe level' message has no use. Rather, they highlight its specific challenges if directly reducing heavier drinking is the goal. Consumers do of course have a ‘right to know’, but how we communicate this information is especially important, as is identifying the goals of any communication campaigns.
The revised Canadian adult guidelines take a more nuanced approached to conveying risk, and it has been suggested that a ‘no risk free level’ of alcohol use might be a better turn of phrase than ‘no safe level’, particularly if it might be less prone to some of the issues around misinterpretation and use. As often with these things, the reality is it’s complicated. Although there may indeed be ‘no safe level’ technically speaking, and some low-risk drinkers do experience harm, there is no one-size-fits-all message for changing attitudes and behaviours.
People drink for many reasons, and increasing their capacity to make informed decisions is important. But when choosing health promotion messages, we must be clear about our objectives and test whether it will achieve them. An intuitively powerful message isn’t always an effective one.
This blog post was written by Dr James Morris, School of Applied Sciences, London South Bank University and Associate for Alcohol Change UK’s Training and Consultancy Unit and Dr Harry Tattan-Birch, Department of Behavioural Science and Health, University College London.
