Prevention of Alcoholism: The Ledermann Model of Consumption (Encyclopedia of Drugs, Alcohol, and Addictive Behavior)
The Ledermann model of alcohol consumption is an important concept for anyone who wishes to understand the underpinnings of modern policy efforts to prevent heavy drinking and alcoholism. The point of departure for this concept is a set of observations about how alcohol consumption is distributed in human societies.
Many have thought of this distribution as occurring in two parts. First, there is the great mass of "normal" drinkers; their drinking might be plotted as a bell-shaped curve, with a few people drinking no more than a sip in a year, an increasing number drinking greater amounts than a sip but less than the average amount, and then a declining number drinking more than the average amount, until the graph reaches the normal drinkers who drink much more than the averagend these are relatively few in number. Second, there is a much smaller number of "abnormal" drinkers; their drinking distribution also might be plotted as a bell-shaped curve, but this curve is shifted to the right of the distribution for normal drinkers. Figure 1 shows this two-distribution concept of normal and abnormal drinking, with the number of drinkers on the y axis and the amount consumed on the x axis.
Sully Ledermann, a French demographer, thought of this problem in relation to a single distribution that was not bell shaped or normal in its distribution. He imagined that drinking ought to be plotted in relation to a single curve, with a shape that is known as "lognormal" and without a categorical distinction between normal and abnormal drinkers. The shape is known as lognormal because the natural logarithms of individual consumption, rather than actual consumption values, are normally distributed. Assuming Ledermann is correct, the majority of individuals within a society will drink relatively modest amount of alcohol, and a small proportion will drink large quantities, but this will appear in an asymmetric or "skewed" distribution curve with a longer tail to the right of the average alcohol-consumption level (see Figure 2). To the right of the curve there should be no bump, which would be caused by the presence of an abnormal-drinkers category, distinct from the category of normal drinkers.
Perhaps the most important implication of Ledermann's thinking about alcohol consumption has to do with the prevention and the reduction of heavy drinking. Categorical distinctions between normal and abnormal drinkers make it possible to focus prevention and intervention efforts on the abnormal drinkers. In contrast, the Ledermann model suggests that efforts can be focused on the great mass of people who drink modestly as well as on the heavier drinkers: In so doing, reductions in the average amount of alcohol consumed should also result in significant reductions in the proportion of people who are very heavy drinkers. This difference in approach is part of an important ongoing debate about how societies can best organize to reduce the hazards of alcohol use.
Ledermann (1915-1967) first proposed his single-distribution hypothesis in a French publication entitled, Alcool, Alcoolisme, Alcoolisation (1956). In a second report published in 1964, he attempted to test and confirm the validity of his theory by using empirical data on drinking behavior from multiple studies. Born in Algeria, Ledermann spent most of his career in Paris, at the National Institute of Demographic Studies (INED) and the University of Paris. A prolific researcher, his interest in the distribution of alcohol consumption within societies developed out of a broader effort to identify the reasons for the lower average longevity of the people in France, in comparison to that of the people in other European countries. Increasingly, he came to believe that a close connection existed between the average, or per capita, level of alcohol consumption within a society and the prevalence of excessive drinkers at risk for alcohol-related injury or death, and that this relationship could be described mathematically.
Ledermann argued that the lognormal distribution of alcohol consumption resulted from the tendency of individuals to develop and change their drinking habits according to a "boule de neige" (snowball) mechanism driven by social pressures. The Norwegian scientist Ole-Jorgen Skog noted that, in general, lognormal distributions tended to result from the exponential (multiplicative) combination of behaviors (1985). On an individual level, this means that persons will tend to increase or decrease their frequency of a behavior by an amount proportional to the initial frequency with which they perform it. For example, we might expect that a person currently consuming 30 liters of alcohol per year would perceive an increase of 6 liters as being comparable to an increase of 1 liter by an individual who currently consumes 5 liters. Such phenomena grow exponentially, in snowball fashion, and tend to distribute according to a lognormal function within populations. Ledermann believed that the snowball effect was caused by the operation of social pressures within drinking environments. This notion implies that the drinking behaviors of individuals within a particular social environment or "drinking culture" are tightly interrelated, such that changes in the alcohol consumption level of some individuals are very likely to induce changes in the consumption level of others. Skog and other scholars have elaborated upon this rudimentary social-interaction hypothesis in an effort to understand how shifts in the drinking habits of one sector may rapidly diffuse throughout the entire population.
The Ledermann model provides a simple formula for estimating the distribution of alcohol use in any homogenous population of drinkers (that is, any population in which the average consumption level does not vary significantly across subgroups). In addition to assuming lognormality with his model, Ledermann also hypothesized that the proportion of drinkers consuming more than 365 liters of absolute alcohol (ethanol) annually was small and invariant across populations, because such high consumption levels (1 liter per day) would quickly have lethal effects. With this constant determined, he could establish mathematically the full distribution of alcohol consumption within a population, knowing only the per capita or average consumption level. Knowledge of the distribution of alcohol consumption yields three important additional insights. First, one can estimate the proportion of heavy or excessive alcohol users in the population. This value is frequently defined as the percentage of drinkers consuming 10 centiliters or more of absolute alcohol per day. Second, the total amount of alcohol consumed by heavy users can be estimated. Third, and most important, the effect of changes in average consumption on the proportion of excessive drinkers in the population can be predicted. This final corollary of the model is perhaps the most controversial, because it indicates that the prevalence of excessive alcohol use within a society can be manipulated by restrictions on alcohol availability or other preventive efforts designed to reduce the general level of consumption in the population. The implications of the Ledermann model for alcohol-control policy and other public health efforts were carefully elucidated in a monograph by Finnish scholar Kettil Bruun and an international body of colleagues (1975).
Ledermann's hypotheses have been the object of intense scrutiny and debate in the half century since they were first proposed. Many researchers have examined the "fit" between the lognormal distribution and data obtained from actual populations of drinkers, with mixed results. Significant deviations from expectations of the model have been demonstrated in some cases; in other populations, the distribution closely approximated lognormality. Ledermann's assumption of constancy across populations of the proportion of heavy drinkers who consume 365 liters or more of alcohol annually has been severely challenged. In general, these critiques have weakened the deterministic character of Ledermann's original formulation, without challenging the basic assertion that there is a close connection between average alcohol consumption in the population and the prevalence of excessive or "at risk" drinkers. The debate over these issues is unresolved, but it is clear that Ledermann's ideas have served as a major stimulus in the effort to understand the relationship between the "drinking culture" of a society and the prevalence of excessive alcohol use.
Ledermann's thinking directly or indirectly underlies many current alcohol policies, especially those that control where, when, and how alcohol is consumed, and how much we pay for it. However, in the half century since his single distribution theory was first proposed, alcohol problem prevention research has continued to grow in sophistication, and modern efforts reflect a greater appreciation of the complexity of societal drinking patterns (Holder et al., 1999; Toomey and Wagenaar, 1999). The assumption of societal homogeneity in drinking behavior was a major tenet of Ledermann's first conceptualization of the single distribution theory. We now have a much greater understanding of the magnitude and significance of variation in drinking behavior, both within and between societies, based on age, gender, ethnicity, locale, and other aspects of culture (Holder and Reynolds, 1998). In addition to level of consumption, alcohol problem prevention efforts also focus on the pattern of drinking and the physical environments where alcohol is consumed. Particular attention in both alcohol and drug abuse prevention studies has centered on such "harm reduction" efforts. This approach focuses on the promotion of safer use patterns rather than limitations on availability (Giesbrecht, 1999; Mosher, 1999). Alcohol server intervention programs, other alcohol education efforts, and early problem identification and intervention programs are examples of this targeted prevention approach. Ledermann's stature and influence in the field of alcohol problem prevention research are still marked, but modern alcohol problem prevention efforts are highly diverse and include a mix of individual and group-based strategies, recognizing that some approaches are appropriately directed at the societal level, but special populations and settings may require focused, specific efforts.
(SEE ALSO: Addiction: Concepts and Definitions; Advertising and the Alcohol Industry; ; Legal Regulation of Drugs and Alcohol; Prevention; Social Costs of Alcohol and Drug Abuse)
BRUUN, K., ET AL. (1975). Alcohol control policies in public health perspective (Vol. 25). Helsinki: Finnish Foundation for Alcohol Studies.
DUFFY, J. C. (1986). The distribution of alcohol consumption0 years on. British Journal of Addiction, 81, 735-741.
GIESBRECHT, N. (1999). Reducing risks associated with drinking among young adults: Promoting knowledge-based perspectives and harm reduction strategies. Addiction, 90, 353-5.
HOLDER, H. D.; FLAY, B.; HOWARD, J.; BOYD, G.; VOAS, R.; GROSSMAN, M. (1999). Phases of alcohol problem prevention research. Alcoholism: Clinical and Experimental Research, 23, 183-94.
HOLDER, H. D., AND REYNOLDS, R. Science and alcohol policy at the local level: A respectful partnership. Addiction, 93, 1467-73.
LEDERMANN, S. (1956). Alcool, alcoolisme, alcoolisation:Données scientifiques de caractère physiologique, economique et social (Institut National d'Etudes Démographiques, Trav. et Doc., Cah No. 29). Paris: Presses Universitaires de France.
LEDERMANN, S. (1964). Alcool, alcoolisme, alcoolisation: Mortalité, morbidité, accidents du travail (Institut National d'Etudes Demographiques, Trav. Et Doc., Cah. No. 41). Paris: Presses Universitaires de France.
SKOG, O-J. (1985). The collectivity of drinking cultures: A theory of the distribution of alcohol consumption. British Journal of Addiction, 80, 83-99.
TOOMEY, T. L., AND WAGENAAR, A. C. (1999). Policy options for prevention: The case of alcohol. Journal of Public Health Policy, 20, 192-213.
BRYAN M. JOHNSTONE