In our recently-published study into alcohol outlets and health in Scotland we found strong correlations between the two: neighbourhoods with higher availability of outlets had higher rates of alcohol-related deaths and hospitalisations. In fact, residents of neighbourhoods with the highest availability were more than twice as likely to die a drink-related death than those with the fewest outlets, all else* being equal (*deprivation and urban/rural status).
But what does this actually mean? Will alcohol-related health problems decrease in a neighbourhood if the number of alcohol outlets decreases? Is the evidence strong enough to advocate reducing outlet availability on public health grounds? These are important questions to ask to ensure that the evidence is used appropriately. Our short answer is that in isolation our study is not definitive, but in combination with other evidence it strongly indicates that reducing alcohol outlet availability will provide health benefits for the Scottish population. To use an analogy, few criminal cases are decided on a single piece of evidence.
In this blog we describe the strengths of our study, and then address the criticisms that it has received. Firstly, what makes this study worth taking notice of?
- Our sample was huge: the 5 million residents of Scotland. This ensures that the results are robust, and applicable Scotland-wide. Also, this was the first population-wide study of alcohol outlets and health in Scotland.
- The relationship is plausible, strong and shows ‘dose-response’: These are widely accepted criteria against which to assess whether an association is likely to be causal (the Bradford-Hill criteria). ‘Dose-response’ means that each change in exposure to the proposed risk factor (here: alcohol outlet density) corresponds with a change in the rate of the health outcome (see graph below for alcohol-related death rates).
- And most crucially: It is the latest piece of independent research in a large number of studies to converge on the same conclusion. An independent review1 of 39 international longitudinal studies of how drinking and related harms change when outlet densities change (e.g., as a result of changes in regulations) concluded that drinking and harms increase with outlet densities. A further 51 correlational studies at a single point in time (i.e., similar to our study) were also included in the review, and more than three-quarters of these supported the main finding. The researchers concluded “…the scientific evidence reviewed indicates that the regulation of alcohol outlet density can be an effective means of controlling excessive alcohol consumption and related harms…” (page 567). Hence, there is a growing body of literature pointing to a causal link, and our findings are consistent with the expected causal pattern.
So our research is robust and consistent, but some strong criticisms have been levelled against it:
- The study ‘failed to find’ a causal link: We did not set out to find a causal link – an impossibility with the available data – but rather to add to existing evidence and extend it to Scotland. To provide the most robust evidence of whether a causal link exists – or not – we invite Scotland’s alcohol retail industry to collaborate with us in a randomised controlled trial. By closing outlets in randomly selected areas, monitoring alcohol-related outcomes before and during the closure (for many years, to capture deaths due to chronic illness such as liver disease), and comparing these with similar outcomes in unaffected ‘control’ areas, we could provide the definitive evidence that critics of the study are seeking. In reality, however, we do not expect such a collaboration to occur.
- It’s just a case of supply and demand: Instead of higher alcohol outlet densities causing increased alcohol consumption it is argued by others that new outlets open in areas where there is a perceived demand for alcohol, or that individuals prone to harmful drinking move into areas with high outlet availability by choice. But in either case alcohol outlets are still supporting drinking, and a wholesale reduction in outlet numbers could reduce alcohol-related harms, via reducing competition, convenience, social aggregation and exposure to marketing, as well as increasing prices and altering existing norms.
- The evidence is not perfect: As acknowledged in the report we considered only a single point in time (a necessity with the available data). A centrally-maintained and archived register of alcohol outlet licenses would assist in conducting longitudinal work in future, but this currently does not exist. But can we afford to wait for ‘perfect’ evidence before taking a course of action that a growing body of work indicates could reduce alcohol-related harms in Scotland? As described above, this would be a long wait. Failure to act on the good evidence available so far poses a danger to public health, and a substantial cost to the taxpayer. With one in every twenty deaths and hospital episodes in Scotland being attributable to alcohol, and with the negative consequences of alcohol costing Scottish society £3.6 billion a year, it would be negligent not to act on the existing evidence.
Reference: 1 Campbell, C.A. et al. (2009) The Effectiveness of Limiting Alcohol Outlet Density As a Means of Reducing Excessive Alcohol Consumption and Alcohol-Related Harms. American Journal of Preventive Medicine 37(6):556-569