Two Dimensions of Alcohol Consumption. Methodological Issues

There are two dimensions of alcohol consumption: average volume of consumption and patterns of drinking, both of which are related to disease burden (Rehm et al., 2003). Average volume, or per capita consumption, is related mostly to long-term health consequences, including alcohol dependence.

Although average volume is also related to acute consequences, such as injuries, several studies indicate that the ability to predict such injuries is increased by taking patterns of drinking into account (Rehm et al., 1996). For example, the same average volume of consumption (1 drink per day) can be consumed in one day of the week (7 drinks on one occasion) or through daily drinking (1 drink per day over 7 days), and the expected outcomes are different. In other words, how an individual drinks, which is influenced by cultural context, can moderate the impact of average volume of consumption on mortality and morbidity.

The pattern of drinking (how, when, and how much is consumed) has been related to acute health outcomes such as injuries and also to chronic diseases such as coronary heart disease and sudden cardiac death (Rehm et al, 2006).

As part of the World Health Organization (WHO) Comparative Assessment of Risk Factors for the Global Burden of Disease study (2000), both dimensions of alcohol consumption for different regions of the world were quantified. It is beyond the scope of this article to cover details of such study, which can be found in other publications (Rehm etal., 2003a, b). This article provides an overview of some methodological issues regarding these two dimensions relevant for comparative studies at the national level, and presents an international trend analysis as well.

Methodological Issues. The WHO has recently updated international guidelines for monitoring alcohol consumption and related harm (WHO, 2000). This guide provides details on how to calculate per capita consumption and related harms, including limitations and problems in interpreting the data. It should be consulted when planning to collect and use national and international estimates.

Estimating Per Capita Consumption by Country. Estimating the average volume of alcohol consumption in a country is best made using national sales and production and/or taxation data, as population surveys invariably underestimate total alcohol consumption. Retail sales data are the most accurate means of estimating how alcohol is consumed in a population, as governments often monitor sales data for tax collection purposes.

Although there are some limitations to this type of data (e.g., beverages can be purchased yet not consumed in the same year, stockpiling can occur before a tax increase, neither home production nor smuggling is accounted for), it is still a relatively reliable source.

When gathering these data, the three major categories of alcoholic beverages (beer, wine, and distilled spirits) available within a country should be included in the estimates. In many developing countries, however, local beverages are as important as the three major categories (e.g., cider, fruit wines, shochu, aguardiente, cachaca, samsu) but often do not fall into the other categories or are sold in informal markets or are not taxed.

Therefore, additional survey data can provide information on who drinks what type of beverages (at least by gender and age groups), which can then be useful in monitoring trends in consumption and relating specific beverages to specific harms.

Per capita consumption figures are given in liters of pure alcohol, which require estimates and/or assumptions about alcohol content of different beverages. Beer, for example, is usually estimated at 5% pure alcohol, but it can vary from 0.9 to 12%. As there is no international standard, countries should make periodic efforts to estimate median alcohol content of each beverage category as they can vary widely from country to country as well as within regions of the same country.

According to WHO recommendations, the total estimated adult alcohol consumption for a country is equal to the total alcohol production plus alcohol imports minus alcohol exports (in the same year) divided by the total population of 15 years of age and over. However, in calculating the average consumption of the total population there may be an underestimate of consumption, particularly in developing countries in which a large proportion of the population is under 15 years of age.

Market research firms serving the alcoholic beverage industry, or industry associations are a good source of data (e.g., Impact Databank, World Drink Trends, Statistical Handbook); however, these data are expensive to obtain and therefore of limited use by governments or researchers. These publications do not cover all countries, especially developing countries, and the information is often not reliable. However, in the absence of any other more reliable source of information, they can be used to supplement data at country and international levels.

The impact of tourism can also be substantial, and some estimate of the size of the tourist population in a given year (15 years of age and older) can be used for estimating adult per capita consumption for a country. If there are special sales taxes or measures for alcoholic beverages sold to tourists that can distinguish them from local consumption, the estimates can be done more adequately.

The United Nations Food and Agriculture Organization (FAO) publishes the most complete set of statistics available on the production and trade of beer, wine, and distilled spirits, and this set can be used in the absence of sales data. These statistics are collected from annual questionnaires sent to ministries of agriculture and trade (they are also available to the public on the FAO’s website, see under ‘Relevant Websites’ section).

The data consist of estimates of production and trade in metric tons for wine; vermouth; must of grape; fermented beverages; spirits; sorghum, millet, maize, and barley beer; and wheat- fermented and rice-fermented beverages. Beverage data are converted into metric tons of pure alcohol and then all beers are combined into a single beer category, all wines into a wine category, and all spirits into a ‘beverages, alcoholic’ or distilled spirits category (WHO, 2000). However, the data rely on national reporting (and not all countries are included) and do not include estimates of unrecorded consumption (e.g., home production, dutyfree alcohol, smuggling).

Nonetheless, the FAO remains the most comprehensive international data source on production of alcoholic beverages. Caution needs to be taken, however, in interpreting the data. Because only large-scale industrial production is quantified, data for countries with substantial informal, low-technology, or home production will be underestimated, as is the case with several African countries. In addition, when data are missing, the gaps are filled with repeated information of the last year reported. As the country population grows, per capita consumption may artificially decline, thus providing a false trend (WHO, 2000).