Wednesday February 27 2013
More than half of all adults may binge drink regularly
"England is a nation of secret boozers," The Independent argues, as it reports on a study investigating the discrepancy between alcohol sales in England and the amount people say they drink in surveys.
International estimates suggest people can underestimate their alcohol intake by around 40 to 60%. In what amounts to be a series of educated guesses based on alcohol sales figures, the researchers came up with new estimates of alcohol consumption by assuming that all drinkers were underestimating their consumption by 40%. They used these hypothetical figures to 'bump up' previous real-life estimates compiled by health surveys.
Using this approach, the researchers estimate that the proportion of adults estimated to be binge drinkers in England increased:
- by 20% in men, pushing the overall estimate up to 52%
- by 28% in women, pushing the overall estimate up to 56%
As the authors admit, assuming that everyone underestimates their alcohol consumption by 40% was a bit of a blunt approach. Also, the difference between sales and reported consumption may be due to many reasons other than under-reporting.
Still, this study serves to highlight that survey data alone is not providing the full picture about alcohol consumption in England. All of us need to be aware that the data from these surveys could underestimate the amount of alcohol people consume and what this means for public health.
Where did the story come from?
The study was carried out by researchers from the University College London Department of Epidemiology and Public Health and was funded by a Medical Research Council Doctoral Training Grant in support of the lead author. No conflicts of interest were declared.
It was published in the European Journal of Public Health, a well-established peer-reviewed journal.
While the general coverage of the study was accurate, some of the reporting ventured into the speculative. A lot of the coverage pushed the idea that British people were 'secret boozers'.
Although deliberate under-reporting of alcohol consumption is probably a factor (possibly due to embarrassment), it cannot be proven by this study. There are probably a number of other reasons why people genuinely underestimate their drinking.
The Daily Express headline "Now 80% of women are 'binge drinking' " is both speculative and inaccurate. The figure appears to come from the estimate that 80% of women (and 75% of men) would exceed the recommended daily maximum of two to three units (three to four for men) on their heaviest drinking day of the week. This is not a binge, which is defined as drinking more than double the recommended daily maximum (six or more units for women, eight for men). Binge drinking had an estimated prevalence of just over half of men and women.
What kind of research was this?
The study aimed to predict the implications of the under-reporting of alcohol consumption in England for consumption above the drinking thresholds recommended by the government.
The researchers state how reported alcohol consumption from surveys typically only accounts for around 40-60% of total alcohol sales conducted internationally, and this is also likely to be the case in England.
Since 1995, the UK Chief Medical Officer has recommended that we do not regularly exceed daily limits of three to four alcohol units a day for men and two to three units a day for women. The Department of Health's definition of binge drinking is consuming more than double the recommended limits in one session: that is, eight units or more for men and six or more units for women.
What did the research involve?
The researchers used results from two nationally representative samples of private households in England to get estimates of self-reported alcohol consumption. These were the General Lifestyle Survey (GLF) and the Health Survey for England (HSE) 2008. Both surveys aim to give representative views of adults aged 16 years and over living in private households in England.
Researchers also obtained data on alcohol sales in England, which suggested higher alcohol consumption than reported by the surveys. The researchers used three different scenarios to account for the under-reporting of alcohol consumption (the difference between self-reported consumption and sales of alcohol). The scenarios were:
- assume equal under-reporting for everyone (under-report by 40%)
- assume under-reporting varies by alcohol consumption level (those consuming more underestimated more)
- assume under-reporting varies by drink type (for example, some people may not consider having a glass of wine to be 'proper drinking')
The second and third were based on factors known to affect the reporting of alcohol consumption.
The impact of these hypothetical scenarios was studied in relation to their impact on:
- the prevalence of drinking more than the UK government weekly guidelines – 21 alcohol units per week for men and 14 for women
- the prevalence of drinking more than the UK government daily guidelines – three to four units per day for men and two to three for women
- the prevalence of binge drinking – defined as consuming eight units or more in one session for men and six or more units for women
The researchers identified many areas that could account for alcohol under-reporting in national surveys, including:
- drinking in those under 16 years of age
- drinking in those outside the survey sample, such as homeless people or people living in institutions, such as people in the armed forces or residential care
- drinkers who simply don't respond to surveys
- alcohol that is bought but not consumed, such as wine that is stored, plus spillage and wastage
- alcohol consumed in the UK by foreign visitors
They carried out a statistical analysis in order to estimate the likely new average alcohol consumption, having adjusted for the under-reporting.
They then estimated how many more people this would push into the binge drinking category, or those exceeding the recommended daily or weekly threshold for alcohol consumption.
What were the basic results?
Average weekly alcohol consumption was available for 12,490 adults in the GLF 2008, and data on the heaviest drinking day in the last week was available for 9,608 adults in the HSE 2008.
After adjusting for an equal rate of under-reporting (scenario 1), average weekly units reported in the GLF survey 2008 increased from 17.1 to 28.0 units in men and from 8.7 to 14.1 units in women.
After adjusting for under-reporting (assuming equal under-reporting for everyone, scenario 1):
- the estimates of the prevalence of drinking more than the weekly guidelines from the surveys increased by 15% for men and 11% in women, such that 44% of men and 31% of women were estimated to drink more than the weekly government guidelines overall
- the prevalence of exceeding the daily limit increased by 19% in men and 26% in women after adjusting for under-reporting, to the extent that 75% of men and 80% of women would have exceeded the recommended limit on their heaviest drinking day in the last week
- the prevalence of binge drinking increased by 20% in men and 28% in women, pushing the overall estimate up to 52% and 56% respectively
The researchers said that the other two hypothetical scenarios gave similar results, but they did not report these findings in detail.
The revision changes some of the significant predictors of drinking above thresholds. In the revised scenario, women have similar odds to men for binge drinking and higher odds of drinking more than the daily limits, compared with lower odds in the original survey.
How did the researchers interpret the results?
The researchers concluded that, "Revising alcohol consumption assuming equal under-reporting across the population does not have an equal effect on the proportion of adults drinking above weekly or daily thresholds. It is crucial that further research explores the population distribution of under-reporting."
This research explored the difference between the amount of alcohol sold in England and that reported to be consumed by national surveys. Modelling the situation in which all alcohol sold in England was consumed in England, they found significant increases in the proportion of men and women exceeding the daily and weekly limits for safe alcohol consumption.
The adjustment significantly increases in the proportion of binge drinkers. They also found that this adjustment changed the pattern of those most at risk of exceeding weekly and daily limits, as well as groups most at risk of binge drinking.
This intriguing research suggests that adults in England are actually drinking much more than normally assumed from survey results alone. While this may well be the case, there are certain limitations of this study to be aware of.
As mentioned, there are many legitimate reasons why levels of alcohol consumption reported in surveys may be different from data on alcohol sales in England, aside from under-reporting. The study assumed that all of the differences between sales data and survey reporting was due to under-reporting. This may not be the case and would over-estimate alcohol consumption levels in England.
However, there is also uncertainty around the exact amount of alcohol sold in England, which the researchers say may be an underestimate. Both scenarios introduce a certain amount of error into the precise estimates of alcohol consumption in England.
In scenario 1, researchers assumed that everyone under-reported their alcohol consumption by 40%. This is likely to be an over-simplification, and the real picture is likely to be more complex and variable between groups.
The researchers highlight the need for much more robust information about national alcohol consumption from many sources, including alcohol sales and surveys of consumption.
If certain groups are under-reporting their alcohol consumption, more research needs to be done in order to firstly identify which groups and why.
Knowing how much alcohol is consumed is key to public health initiatives aimed at reducing alcohol consumption to within healthy limits.
This study usefully serves to highlight the important point that survey data alone cannot provide the full picture on alcohol consumption, or other issues. Many people have a tendency to tell health professionals what they think they want to hear rather than the complete truth, an issue that this study did not address.
People working in healthcare – as well as the general public – need to be aware that alcohol consumption can be underestimated and of the potential impact this could have on public health.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.