Perceived academic performance and alcohol, tobacco and marijuana use: Longitudinal relationships in young community adolescents

Abstract

Objective

To study longitudinal relationships between perception of academic performance (PAP) and alcohol, tobacco and marijuana use at age 13 and 15 years.

Method

Self-report data from students in 27 schools in South Australia (n = 1579) was analysed with ANCOVA and logistic regression, controlling for confounding factors (socio-demographic, parenting, depressive symptoms, anxiety, self-esteem, antisocial behaviour).

Results

Persistent or increasing perception of academic ‘failure’ (self-rated failing or below average performance), compared to improving or stable perception of average (or above) performance, from age 13 to 15 years, predicts more than weekly alcohol and tobacco use at age 15 (3- to 4-fold risks), in addition to increased risks from early substance use (e.g., 12- to 15-fold for triple combinations), controlling for confounders. Increased risks for more than weekly marijuana use at age 15 are 3-fold in those with increasing perception of failure; though relationships are fully mediated by antisocial behaviour in those with persistent perceptions of failure.

Conclusion

Interventions to maintain or improve academic self-esteem in early adolescence may reduce risks for accelerating substance use in mid-adolescence.

Introduction

Substance use in adolescents is a widespread and potentially hazardous activity, increasing risks for dependence and abuse, and other adverse physical and psychosocial outcomes (Lewinsohn, Rohde, & Brown, 1999). Although US Monitoring the Future 2003 data suggest that alcohol, tobacco and marijuana use is in decline for older adolescents, the trend is less clear among 8th graders, perhaps reflecting a generational shift (Johnston, O'Malley, Bachman, & Schulenberg, 2004). Further, age of initiation is becoming younger (Degenhardt, Lynskey, & Hall, 2000); and early onset drug use is associated with both greater likelihood of continuation (Chen & Kandel, 1998) and greater long-term harm and dependence (Grant & Dawson, 1997).

Not all substance users in adolescence progress to problematic levels, and psychosocial risk models describe a complex interplay between individual and personality factors, environmental social and family factors, and drug-taking behaviour (Jones & Heaven, 1998). Twin studies show that environmental and social rather than genetic factors influence substance initiation and use (Han et al., 1999, Lynskey et al., 2003). Specifically, known risk factors include antisocial behaviour and substance using or delinquent peers (e.g., Hops, Davis, & Lewin, 1999); depressive symptomatology (Rohde, Lewinsohn, & Seeley, 1996); family functioning, parenting style and expectancies (Cohen & Rice, 1997); and childhood abuse (Bergen, Martin, Richardson, Allison, & Roeger, 2004). Substance use may also increase risks for other drug-taking. Gateway theory describes the increased likelihood of progression from one substance to another, though the exact sequencing and causal attribution to associations are issues still in debate (Kandel, 2002).

Substance use as early as grade 6 or 7 is associated with, and possibly causes, adverse outcomes in academic motivation and later achievement (Ellickson et al., 2001, Jeynes, 2002). Conversely, poor academic achievement in grades 6 and 7, as well as problem behaviour, and low engagement with and negative perceptions of school, predict later alcohol initiation and misuse (Diego et al., 2003, Sobeck et al., 2000). Global and academic self-esteem from age 9 to 13 (McGee & Williams, 2000) and academic and social behaviour at age 7–9 (Hops et al., 1999) predict health compromising behaviour and substance use at age 15. In contrast, high engagement with school is protective against alcohol misuse (Aunola, Stattin, & Nurmi, 2000). Structural equation modelling of relations between cigarette use and school factors from 8th to 12th grade, suggest that the direction of influence is from school experience to cigarette use (Bryant, Schulenberg, Bachman, O'Malley, & Johnston, 2000).

Rarely explored, however, is that general societal pressure to succeed, high parental expectation of academic achievement, and associated perception of failure may contribute to ineffective coping mechanisms such as substance use. A report on Canadian female adolescent health risk behaviour raises the question that schools may contribute to these problems, and that “a supportive school environment and less emphasis on academic achievement are needed” (King, 1998). Further, Kumpulainen and Roine (2002) found that early (age 12) perception of school failure and low self-esteem in girls, and interpersonal problems and aggressive tendencies in boys, were more important predictors of heavy alcohol use at age 15 than early depressive symptomatology. Sutherland and Shepherd (2001) found strong relationships between social factors and substance use outcomes in 11–16 year olds (n = 4516), with perceived academic performance ranking third in importance after ‘concurrent second or third substance use’, and ‘having been in trouble with police’ as predictors of substance use.

This investigation explores new directions in associations between perceived academic performance (PAP) and substance use, extending the study of Kumpulainen and Roine (2002), who appealed for more investigation into academic expectations and alcohol use in young adolescents. First, PAP is a psychosocial construct related to school, family and individual expectations, academic self-esteem and current mood. It may or may not be related to actual achievement, thus possibly encompassing cognitive distortions as well. We construct PAP trajectories from assessments at age 13 and 15, categorising perception of academic performance broadly as one of ‘failing at both times’, ‘average at both times’, ‘improving from failing to average’, or ‘declining from average to failing’; this enables longitudinal tracking over the crucial 13–15 years period.

Second, we consider combinations of alcohol, tobacco and marijuana use, enabling comparison of abstainers at age 13 with users of single, double or triple combination(s) at age 13, in regard to associations with more than weekly substance use at age 15. This ensures control for early substance use, which is important and sometimes not taken account of in other studies.

Third, to elucidate any unique contribution that PAP may make to substance use at age 15, we control for other known risk factors such as antisocial behaviour, depressive symptoms and anxiety, family functioning, parenting style and family living arrangement.