Are traditional measures of alcohol use appropriate worldwide?
In any research on the impact of alcohol on health and well-being it is important to have good measures of how, what, and how much people drink. Last month (23rd September 2016) we held the second meeting of the Alcohol Research Interest Group at the London School of Hygiene & Tropical Medicine (LSHTM) on the theme “Measurement of Alcohol Use”. This meeting was part of a joint ongoing initiative to bring together LSHTM researchers with a shared interest in alcohol.
What was discussed during this meeting?
There were two presentations: The first presentation given by David Leon was on Alcohol Biomarkers, their strengths and limitations, with some examples of their use from his own work in Russia. Alcohol biomarkers are physiological parameters which can be measured for example in blood, urine or hair which provide an indication of exposure to alcohol. These could either be related to very acute exposure with applications to forensics (identifying drunk drivers, alcohol levels at autopsy) or more long term exposure such as markers of alcohol-related liver damage. The attraction of using alcohol biomarkers in research is to provide more objective measures of alcohol compared to relying of self-reported data about how much people drink, which is likely to be inaccurate due both to difficulty with accurate recall and perceived stigma which means alcohol consumption is commonly under-reported. While the number of alcohol biomarkers available is expanding, a common limitation is that they cannot differentiate well between different long term drinking patterns such as episodic “binge drinking” compared to frequent heavy consumption. However there are potential benefits to using self-reported data and alcohol biomarkers alongside each other for “triangulation” purposes.
The second presentation by Mia Crampin on the collection of data on alcohol use in Malawi was sadly cut short by technical problems with Skype, but we looked through her slides as a group and they sparked an interesting discussion about the challenges of measuring alcohol use in settings where standard questionnaires, such as the widely used Alcohol Use Disorders Identification Test (AUDIT), may not be appropriate.
Mia’s slides showed pictures of some types of alcohol consumed in Malawi, including homemade alcohol and alcoholic drinks sold in sachets and cartons. Discussing this with her afterwards, the reason for showing the picture of the sachets was their high market penetrance since these are a very cheap source of alcohol which can be purchased by children. Without knowing what would be said about these pictures, we ended up having a discussion from our own experiences about challenges of measuring the amount consumed from sources of alcohol (such as homemade beverages) where it is hard to define what a “portion” is and alcohol content (strength) is variable, and often extremely high. A conventional measurement approach for alcohol consumption relies on asking people questions about how often they drink alcohol (frequency) and the usual amount consumed per drinking occasion based on the concept of a “standard drink,” which is a set size of a known strength beverage (for example a 100ml glass of wine, 13% ethanol content). This is then used to estimate a summary measure of the volume of alcohol consumed. This approach was designed for use in countries where alcohol is consumed in relatively standard measures such as bottles of beer or glasses of wine. Several of us within the meeting had experienced difficulties trying to measure alcohol consumption in settings where the concept of a “standard drink” is not well understood or meaningful. Examples mentioned were Uganda, South Africa, India and Russia.
What was the outcome?
At the meeting we agreed on the need for the development of appropriate tools for the measurement of alcohol which take account of the different ways alcohol in consumed globally. An idea raised was to work together to develop a toolkit with alternative approaches for measurement of alcohol use which do not rely on asking people to estimate the number of drinks they consume. Other approaches to consider include asking about short term effects of heavy drinking such as hangover or drunkenness and using alcohol biomarkers alongside self-reported data from questionnaires. This is an area we are keen to follow up on and to hear about the experiences of others.
We would welcome your thoughts in the comments section below.
---
The Alcohol Researchers Interest Group at the London School of Hygiene & Tropical Medicine is a joint initiative between the Centre for Global Non-Communicable Disease, the Centre for Global Mental Health, and ECOHOST (The Centre for Health and Social Change).
Image credit: Sarah McDevitt
