Ever since the 1960s, smoking has been recognized as the most significant cause of preventable morbidity and premature death. In Canada alone the estimated annual excess mortality from smoking is more than 36,000 (Mao, Gibbons, & Wong, 1992). A large number of epidemiologic studies have linked smoking to cancer, coronary heart disease, emphysema, chronic bronchitis, reproductive-related problems, and other conditions (Surgeon General, 1982, 1983, 1984, 1989, 1990). Despite this evidence, however, an alarming proportion (about one third in 1989) of adult Canadians continue to smoke (Mao et al., 1992). Moreover, adolescent smoking has declined very little in the past decade.
In recent years, health professionals have shifted their emphasis from treating adult smokers to preventing smoking among children (McCaul et al., 1982), since almost all adult smokers initiate the habit during adolescence, typically between the ages of 12-14 (Evans, Henderson, Hill, & Raines, 1979). Moreover prevention of smoking among adolescents is considered important not only because of its deleterious health effects, but because evidence suggests that cigarette smoking may act as a gateway for other drug and alcohol use and abuse (Kandel, 1975).
The major public health problem associated with smoking, coupled with the shift to prevention among youth, has prompted a number of studies of the determinants of smoking behavior among adolescents. These studies have identified a range of factors, including poor family relationships, parental influences, peer pressure, and cigarette advertisements (Murphy & Price, 1988; Goddard, 1990; Hover & Rosenthal-Gaffney, 1988). In the 1960s low self-esteem was added to this list (Bynner, 1969). While some recent studies (Bonaguro & Bonaguro, 1987) have associated low self-esteem with cigarette smoking, others have conflicting opinions as to its relationships to self-esteem and health behavior. In a summary on health behaviors among youths, Bruhn and Parcel (1982) stated that there is little consistent evidence to demonstrate that beliefs about the self are highly correlated with behavior. Although the correlation between smoking and self-esteem is not clear, school-based health education programs, aimed at raising adolescents self-esteem, have evolved to prevent smoking among adolescents.
Gender differences in level of self-esteem are well documented in the literature. Although some studies have failed to detect any differences (Mullis, Mullis, & Normandin, 1992; Maccoby & Jacklin, 1974), most researchers agree that females generally have lower self-esteem than do males (Rosenberg & Simmons, 1975; Brack, Orr, & Ingersoll, 1988). Psychosocial approaches to smoking prevention, however, do not assume any gender differences and treat the adolescent population as a whole is attempting to enhance self-esteem. Recent smoking prevention programs assume that a child with low self-esteem at any given time is more likely to smoke than one who has high self-esteem. This approach targets those with low self-esteem on the assumption that it is stable throughout adolescence. There are, however, conflicting findings regarding age differences in self-esteem which reflect the use of different study designs: some longitudinal studies have indicated that self-esteem among adolescents increases over time while some cross-sectional studies (Mullis, Mullis, & Normandin, 1992), have not yielded the same results.
In addition, previous studies that attempt to show an association between smoking and self-esteem among adolescents (Murphy & Price (1988); Bonaguro & Bonaguro, (1987); Penny & Robinson, 1986) have compared mean self-esteem scores between groups of smokers and non-smokers. Using the mean score as a summary statistic of self-esteem for the whole population does not adequately reflect the distribution of scores in that population. Moreover, this summary score is hard to interpret and may simply represent a skewed population in which a large proportion of adolescents score either high or low. …