Smoking dependence in 18 European countries: Hard to maintain the hardening hypothesis

When the prevalence of smoking decreases in a population, there is a hypothesis-the so-called "hardening hypothesis"-that the remaining smokers form a subgroup of "hardcore smokers." Our aims were to test the hardening hypothesis and to analyze the determinants of high dependence taking into account both individual and country-level characteristics.

Within the Pricing Policies and Control of Tobacco in Europe (PPACTE) project, we conducted a face-to-face survey on smoking between January and July 2010 in 18 European countries, including 2882 male and 2254 female smokers with complete information on smoking dependence. The Heaviness of Smoking Index (HSI) was used as a measure of tobacco dependence. We correlated smoking prevalence and dependence using the country as unit of analysis. Moreover, we fitted multilevel logistic regression models.

Country-specific prevalence of smoking was positively, although not significantly, correlated with the proportion of highly tobacco-dependent smokers (overall rsp=0.203, p=0.419), both in men (rsp=0.235, p=0.347) and women (rsp=0.455, p=0.058). Using individual-level analysis, high dependence was positively related to age, and, although not significantly, to smoking prevalence, and inversely related to level of education. The lack of a smoking ban at home was positively related to smoking dependence.

Using both ecological and individual-level analyses, the relations between smoking prevalence and HSI were not significant, but in the opposite direction as compared to that assumed by the "hardening hypothesis." Therefore, our data provide empirical evidence against this theory, thus supporting the feasibility of an endgame strategy.