Measuring alcohol in populations
Better monitoring of alcohol consumption is needed; technological advances may soon provide extremely valuable information about context of drinking, says Dr Annie Britton
28 October 2016 – There are many reasons why we need to accurately measure how much alcohol people are drinking. It is essential for monitoring trends over time and to make country comparisons. We need the measures to evaluate alcohol policies and we need to be able to link alcohol intake to related harm.
So much of what we rely upon is data from large scale surveys and longitudinal studies. Typically, participants in these studies are asked to self-complete information about their alcohol intake. Clearly, this has many limitations. First, there is no worldwide agreement on what questions to ask and how to measure volume, frequency, binging, heavy drinking and dependency. Individuals are expected to be able to self-recall the details about their alcoholic drink consumption. Herein lays further problems. For example, measures poured in licensed premises may vary hugely from measures poured at home and the alcoholic content of drinks has changed over time.
All of which makes us question the validity and reliability of the data we have at our disposal. But what are the alternatives?
Some researchers make use of contrived settings (e.g. fake bars) to maximise confidence in measurement of alcohol intake, but this raises questions of external validity and is of limited value in evaluating drinking over time or across contexts.
Objective alcohol measures exist, such as biochemical measures of alcohol concentrations from breath, blood, urine and hair; each of which has its limitations. Alcohol is rapidly metabolised by the body. Biological assays of alcohol metabolites may provide valid indicators of heavy consumption in recent days, but cannot provide information about the quantity or frequency of drinking episodes. Urine samples can provide results covering a longer period than breathalyser tests, to a maximum of about 72 hours, but like breathalyser results they are also affected by the rate at which an individual metabolises alcohol. Blood testing can detect alcohol between 14 and 28 days after its consumption. Indices include the liver function test (LFT) and the elevation of carbohydrate deficit transferrin (CDT). Both of these indicators are present when an individual has consumed moderate to heavy amounts of alcohol and therefore the results are not helpful in detecting patterns of low use. Plus the results can be abnormal where the individual has liver damage for reasons other than alcohol consumption (e.g. hepatitis). In hair strand testing, two markers, ethyl glucuronide (EtG) and fatty acid ethyl esters (FAEE), are used to identify ethanol within the body up to a period of 6-12 months. But these markers cannot reveal detailed information about pattern of consumption.
None of the above is ideal and given the size of population surveys, they are prohibitively expensive and not feasible to deploy. One potential solution that could possibly be used in large studies is the transdermal alcohol sensor. Recent technology has been developed that measures the small fraction (approximately 1%) of ingested alcohol that is excreted through the skin via sweat glands and diffusion. The sensors can provide a continuous estimate of ingested alcohol over extended periods of time. They are widely used in the United States to monitor sobriety as part of parole conditions, but could be used to monitor daily patterns. At present the device is worn as an ankle bracelet and if combined with a global positional system, could provide extremely valuable information about context of drinking.
One day the issue of accurately measuring alcohol consumption among populations may be taken for granted. New devices are likely to be smaller, smarter and more acceptable (for example, apparently there is already a tattoo that can measure your alcohol intake). Monitoring devices, such as Fitbits, are increasingly common and alongside this new evolving technology will come our confidence in the data upon which we rely.
Written by Dr Annie Britton, Reader in Epidemiology at University College London.
All IAS Blogposts are published with the permission of the author. The views expressed are solely the author's own and do not necessarily represent the views of the Institute of Alcohol Studies.