Alcohol sales linked with alcohol deaths across GB regions
For the first time alcohol sales data, which act as a proxy for alcohol consumption, and alcohol-related deaths for regions in Great Britain have been analyzed. The results, published in the open access journal BMC Public Health, show that of eleven regions analyzed, the South West, Central Scotland, North East, North West and Yorkshire had higher levels of alcohol sales per adult than the GB average. There were lower sales in London, Central England and the East of England.
The study, led by NHS Health Scotland, also found some interesting regional differences in the types of alcohol sold through different alcohol retailers. The high volume of sales in South West England was mostly due to spirits and cider sold through pubs and restaurants, but sales of wine through supermarkets and other off-licences were also high. Both Scottish regions, but particularly Central Scotland, had substantially higher spirits sales than any other region, especially through off-licenses. In Yorkshire and the northern English regions, more beer was sold per adult than other GB regions. The low volume of sales in London was because of lower sales through off-licences.
It was also found that those regions with higher sales per adult typically had higher rates of alcohol-related deaths. However, the relationship was weakened by the South West region, which had the highest consumption level but one of the lowest alcohol death rates. The researchers at NHS Health Scotland and the Glasgow Centre for Population Health suggest that this could be due to tourism. The South West, which for the study consisted mainly of Devon and Cornwall, had one of the smallest populations but also had one of the highest rates of second homes. People who visit this area but do not live there would increase the sales figures for alcohol and would not show up for alcohol-related deaths.
All previous studies have looked at alcohol consumption for regions in Great Britain based on self-reported data, which is often subject to sampling bias, low response rates and recall bias. The use of alcohol sales data is a more accurate and objective way to estimate consumption. This is still subject to its own biases, such as wastage and spillage; consumption by tourists; and unrecorded alcohol, which includes homebrew, smuggled alcohol and alcohol intended for industrial and medical use.
Mark Robinson, Public Health Information Manager at NHS Health Scotland and study lead, said: "Our study provides support for the relationship between alcohol consumption and alcohol-related mortality across regions in Great Britain, which hasn't always been the case using survey data to estimate consumption. Future studies should consider the use of data from a range of different sources to provide a better understanding of alcohol consumption in GB, its relationship with alcohol-related harms, and the impact of different alcohol policy approaches."
Researchers from NHS Health Scotland regularly use alcohol sales data as a way of monitoring and evaluating the Scottish government's strategy on alcohol consumption.
Alcohol sales data was obtained from market research companies for the years 2010 and 2011. These data were combined and averaged to account for random fluctuations. The researchers converted the total volume in litres to pure alcohol - alcohol by volume (ABV) - for seven drink categories: spirits, wine, beer, cider, perry and pre-mixed alcoholic beverages. Consumption per adult was then calculated by dividing the total population aged over 16 years across regions of GB.
Regional alcohol consumption and alcohol-related mortality in Great Britain: novel insights using retail sales data, Mark Robinson, Deborah Shipton, David Walsh, Bruce Whyte and Gerry McCartney, BMC Public Health 2015, 15:1. www.biomedcentral.com/1471-2458/15/1