Gender differences in trajectories of depressive symptomatology and substance use during the transition from adolescence to young adulthood

Abstract

This study examines gender differences in the association between symptoms of depression and substance use during the transition from adolescence to young adulthood. Data are from three waves of the US-based National Longitudinal Study of Adolescent Health (n=10,828). Results from latent growth curve analysis demonstrate that the association between depressive symptomatology and substance use is bi-directional. Adolescents who are initially more depressed begin the study period with substantially higher levels of substance use than their better-adjusted peers, yet they are less vulnerable to increases in smoking (girls only), binge drinking (girls and boys), and illicit drug use (girls only) across the transition to young adulthood. Also, adolescents who start out with higher than average cigarette, alcohol, and illicit drug use experience a faster rate of decline in symptoms of depression over time compared to those who start out with lower levels of substance use. This association appears to be more pronounced for girls than for boys. Despite their faster rate of decline in depressive symptoms, girls and boys who have higher initial levels of substance use report higher levels of depressive symptomatology at all three time points.

Introduction

The transition from adolescence to young adulthood is a developmental period characterized by significant continuity in depressive symptomatology (Garber & Flynn, 2001), as well as increased risk-taking behavior (Arnett, 2000). Symptoms of depression and risk behaviors, such as smoking, binge drinking, and illicit drug use, may feed off each other during this transition, leading to an accumulation of health-related disadvantage over time (Dannefer, 2003). According to the increased heterogeneity model, the transition from adolescence to young adulthood is challenging and may magnify individuals’ existing strengths and weaknesses. As a result, interindividual differences in functioning and adjustment will grow larger over time. Persons experiencing difficulties during adolescence may be less able to make a successful transition to young adulthood and will, therefore, fall even further behind their better-functioning peers (Schulenberg & Maggs, 2002). For example, a depressed adolescent may drink heavily to cope with unpleasant feelings, and heavy drinking may, in turn, lead to increased symptoms of depression in young adulthood.

Previous research on adolescents and young adults in Western societies strongly suggests that there is a link between depression and substance use (Diego, Field, & Sanders, 2003; Fergusson, Horwood, & Swain-Campbell, 2002; Green & Ritter, 2000; Hallfors, Waller, Bauer, Ford, & Halpern, 2005; Kumpulainen, 2000; Martini, Wagner, & Anthony, 2002; Poulin, Hand, Boudreau, & Santor, 2005; Saules et al., 2004; Silberg, Rutter, D'Onofrio, & Eaves, 2003; Waller et al., 2006; Wu & Anthony, 1999). Some researchers argue that depressed persons use substances, such as marijuana or alcohol, to cope with symptoms of depression (see Kumpulainen, 2000), while others contend that substance use actually causes depression (see Wu & Anthony, 1999). Taken together, the results of previous research suggest that the association between depressive symptomatology and substance use may be bi-directional, yet few studies have empirically examined this possibility. Notable exceptions include work by Hallfors et al. (2005) and Wu and Anthony (1999), who find that substance use predicts increased depression, but depression does not predict increased substance use.

This study extends previous research by examining depression and substance use as dynamic, interrelated trajectories, conditioned by gender. Using nationally representative panel data from the United States, this study specifically considers bi-directionality in the association between symptoms of depression and substance use. Consistent with the life course perspective, which states that development occurs over time and within social contexts (Elder, 1998), the present study uses latent growth curve modeling to determine whether trajectories of change in depressive symptomatology are associated with trajectories of change in smoking, binge drinking, and illicit drug use during the transition from adolescence to young adulthood and whether gender, an important aspect of social structural context, influences the association between trajectories of depression and substance use.

Numerous studies find that cigarette use is more common among depressed adolescents (Diego et al., 2003; Martini et al., 2002; Wu & Anthony, 1999) and adults (McChargue, Spring, Cook, & Neumann, 2004; Saules et al., 2004; Ulrich, Meyer, Rumpf, & Hapke, 2004) than among their non-depressed peers. In a longitudinal study of approximately 2000 female students at the University of Michigan, Saules et al. (2004) found that increasing symptoms of depression during the first year of college were associated with late-onset smoking. They speculate that smoking may represent an attempt at self-medication among the young women in their study. Other research suggests that young adults with a history of depression may be more likely to start smoking because they hold exaggerated expectations about the positive effects of nicotine (McChargue et al., 2004). Furthermore, depressed smokers appear to value the pros of smoking more highly than other smokers, making them less likely to quit smoking once they have begun (Tsoh & Hall, 2004).

While many researchers argue that depressed mood leads to the initiation of smoking, others find that cigarette smoking during adolescence may lead to depression (Martini et al., 2002; Wu & Anthony, 1999). In a prospective study of predominantly African-American school children in the mid-Atlantic region of the United States, Wu and Anthony (1999) report that cigarette smoking is associated with a moderately increased risk of depressed mood during the transition from childhood to adolescence; they find no association, however, between early depression and later smoking. Martini et al. (2002) suggest that smoking may be linked to depression through social stigma associated with cigarette use. Based on cross-sectional data from the US National Household Survey on Drug Abuse, they find that current teen smokers have higher levels of depressive symptomatology than former smokers and never smokers. In addition, among former smokers, they report that depression levels are inversely associated with time elapsed since last smoking, offering support for the hypothesis that smoking causes higher levels of depressive symptomatology among adolescents (Martini et al., 2002).

Cross-sectional research suggests that depression and other forms of substance use, including heavy drinking and illicit drug use, are positively associated with one another (Diego et al., 2003; Green & Ritter, 2000). Green and Ritter (2000) contend that drug use may be a type of avoidance coping and that depressed persons may be more likely than their non-depressed peers to use marijuana to cope with problems. They find that when young adult males use marijuana as a coping mechanism, it is associated with greater depression, but when they do not use the drug to cope with problems, it is actually associated with decreased symptoms of depression. They conclude that it is unclear whether using marijuana to cope leads to an increase in depression or whether those who use the drug to cope have more symptoms of depression to begin with (Green & Ritter, 2000). Similarly, Diego et al. (2003) find that depressed adolescents are more likely than others to drink alcohol and use marijuana. They suggest that the relationship between substance use and depression may be bi-directional but do not directly test this hypothesis.

Two longitudinal studies offer differing conclusions regarding the direction of causation between depression and substance use. Results from a study of New Zealand adolescents aged 14–21 suggest that weekly users of marijuana are at increased risk of depression (Fergusson et al., 2002). The authors of this study argue that the regular use of marijuana may have multiple negative consequences, including increased risk-taking behavior, depression, and suicidality (Fergusson et al., 2002). In contrast to these results, a longitudinal study of young people in Finland finds that symptoms of depression at age 12 are positively associated with alcohol use three years later (Kumpulainen, 2000). This researcher concludes that depressed adolescents may find relief from their symptoms in drinking alcohol, and may, therefore, be more eager than their non-depressed peers to consume alcoholic beverages (Kumpulainen, 2000).

Before adolescence, boys tend to have higher rates of depression than girls. By age 13 or 14, however, girls are much more likely than boys to be depressed; and after the age of 15, adolescent girls and adult women across cultures are twice as likely as boys or men to be depressed (Nolen-Hoeksema (1990), Nolen-Hoeksema (2001)). Although adolescent girls and adult women are substantially more likely to suffer from symptoms of depression, their male counterparts are more likely to engage in behaviors, including smoking, heavy drinking, and illicit drug use, that increase their risk of disease and death (Courtenay, 2000a). While these gender differences are well-documented, much less is known about potential variation, within gender groups, in the relationship between depressive symptomatology and substance use.

Some studies suggest that depressed girls and women in Western societies may be more likely to engage in health risk behaviors than depressed boys and men. For example, one longitudinal study of twins reports that depression leads to smoking, alcohol use, and drug use in girls more than in boys (Silberg, Rutter, D'Onofrio, & Eaves, 2003). Furthermore, among people with alcohol use disorders, Nolen-Hoeksema (2004) finds that women are more likely than men to use alcohol to regulate negative affect.

Other evidence suggests that boys and men may be more vulnerable to the consequences of depression for substance use. In a study based on data from Finland, Kumpulainen (2000) finds that depressed boys are more likely to use alcohol than depressed girls. Furthermore, among social drinkers, men are more likely than women to use alcohol to regulate negative affect (Nolen-Hoeksema, 2004). Theoretical work on gender is consistent with these empirical findings, which indicate that the path from depression to substance use may be stronger for boys than girls. Courtenay (2000b) argues that men oftentimes engage in health risk behaviors to demonstrate hegemonic masculinity. Boys and young men may smoke, drink, or use drugs to appear strong, masculine, and in control. For depressed males, who likely sense that they are more weak or vulnerable than peers, engaging in health risk behaviors may be an easily accessible, effective means of enacting masculinity.

Several studies suggest that girls may be more vulnerable to the consequences of substance use for depressive symptomatology. In a cross-sectional study using the Add Health data, Waller et al. (2006) found that girls who engaged in light and moderate risk behavior, including drinking, smoking, and sexual intercourse, experienced significantly more depressive symptoms than boys engaged in similar patterns of behavior. In another recent cross-sectional study of adolescents in Canada, researchers found that alcohol use and cigarette smoking are associated with increased symptoms of depression in girls but not boys, while marijuana use is associated with higher levels of depressive symptomatology for both girls and boys (Poulin et al., 2005). Finally, a longitudinal study using the Add Health data found that experimental and high-risk patterns of sexual behavior and drug use at Wave 1 predicted increased depression for girls, whereas only high-risk patterns of behavior were associated with increased symptoms of depression among boys (Hallfors et al., 2005).

Depression and substance use are important public health problems. According to the World Health Organization Global Burden of Disease Study, depression is the single most burdensome disease in the world in terms of total disability-adjusted life years among the middle-aged (Murray & Lopez, 1996). This top ranking is attributed to a combination of high lifetime prevalence, early age of onset, high chronicity, and high role impairment (Kessler, 2000). Smoking is associated with a broad range of diseases and causes of death, including lung cancer, cardiovascular disease and stroke, emphysema, other forms of cancer, and diabetes, as well as other ailments and social pathologies. In a longitudinal analysis of adults in the US, smoking led to a 100% increase in the odds of death over the follow-up period. Results of this study also suggest that heavy drinking leads to increased risk of cancer and social pathology (Rogers, Hummer, & Nam, 2000).

A better understanding of the relationship between depression and substance use is necessary for improving the health of adolescents and adults in the US and throughout the world. Previous research strongly suggests a bi-directional association between depressive symptomatology and smoking, heavy drinking, and illicit drug use. In order to minimize the negative consequences of depression and substance use, health care providers need to understand how these problems develop over time, which subpopulations are most affected by them, and the extent to which they are interrelated. Thus, the first step in this analysis is to describe trajectories of depression and substance use across the transition from adolescence to young adulthood. Given the well-documented gender differences in depressive symptomatology and substance use, the next step is to examine how these trajectories differ for young women and men. The final step in this analysis is to determine whether trajectories of depression and substance use are related across time and whether there are gender differences in this association.