Relative risks of adolescent and young adult alcohol use: The role of drinking fathers, mothers, siblings, and friends

Abstract

The present study examined to what extent regular drinking of fathers, mothers, co-twins, siblings, and friends was related to adolescent regular drinking in three age groups: 12–15, 16–20 and 21–15-year olds. The sample consisted of 3760 twins (1687 boys, 2073 girls) with a mean age of 17.8 years. Data were based on twins' self-reported alcohol uses and reports about siblings' and friends' alcohol use, and on parents' self-reports. Results showed that generally in each of the three age groups, regular drinking of same-sex co-twins and friends posed the highest risk for regular drinking. Age differences indicated that these risks decreased with age. Irrespective of age, regular drinking of fathers and mothers posed the lowest risk. Findings were generally the same for males and females.

Introduction

Alcohol misuse among adolescents is a substantial problem throughout the Western world. Recent figures indicate that experimentation with alcohol is rather normative in adolescence and that many of the adolescents develop a regular drinking pattern (Hibell et al., 2004, Poelen et al., 2005, Trimbos, 2004). For example, 85% of the secondary school pupils have experimented with alcohol and 58% had used alcohol in the previous month in 2003 in The Netherlands (Trimbos, 2004). Since adolescent alcohol use has been related to a wide range of problems such as school problems and aggression (Gruber et al., 1996, Wechsler et al., 1994), alcohol related injury and deaths (Hingson, Heeren, Zakocs, Winter, & Wechsler, 2003), suicidal ideation (Light, Grube, Madden, & Gover, 2003) and even impaired brain development (Tapert, Caldwell, & Burke, 2004), ample research has been devoted to exploring the correlates of adolescent alcohol use. One substantial factor that has been identified is the drinking behavior of persons in the adolescents' direct social environment, most notably the peer group and the family. Various studies have examined the effects of parents' and friends' drinking, but have not yet provided a conclusive overview of the relative impact of the alcohol use of each of these persons. The aim of the present study was, therefore, to examine the relative risk of adolescent alcohol use by comparing the associations of adolescent alcohol use with fathers', mothers', siblings', and friends' use within a single sample.

The direct link between parental drinking and adolescent drinking seems far from clear. A number of studies revealed that parental alcohol use is related to adolescent use (e.g., Ary et al., 1993, Ellickson and Hays, 1991, Hawkins et al., 1997, Van der Vorst et al., 2005, Webb and Baer, 1995, Webb et al., 1991), and that adolescents who had heavy drinking parents were more likely to drink heavy as well (Cohen & Rice, 1997). In contrast, however, other studies found no direct links between parental drinking and adolescent drinking (Boyle et al., 2001, Ouellette et al., 1999, Petersen et al., 1994, Power et al., 2005, Reifman et al., 1998).

A vast body of research has documented that a strong association exists between friends' and adolescents' alcohol use (Andrews et al., 2002, Petraitis et al., 1995, Sieving et al., 2000, Thombs et al., 1997). For example, friends' drinking predicted adolescents' alcohol use initiation and cessation over a one-year period (Maxwell, 2002). Bot, Engels, Knibbe, and Meeus (2005) reported that when an adolescent's best friend was drinking intensively, it was likely that the adolescent also drank intensively one year later, irrespective of whether the friendship was stable across this one-year period. Likewise, perceptions of the number of drinking friends have found to be predictive of adolescents' later alcohol consumption (Ouellette et al., 1999, Simons-Morton and Chen, 2006). Thus, it appears that friends' alcohol use constitutes a significant risk factor for adolescent drinking, although recent studies suggest that it is peer selection rather than peer influence that contributes to similarity in health risk behaviors (Jaccard, Blanton, & Dodge, 2005).

Siblings' alcohol use may constitute another potential risk factor for adolescents' alcohol use. However, the role of siblings in relation to adolescent drinking has been explored only in a few studies. These studies report direct associations between siblings' and adolescents' substance use in general (Brook et al., 2003, Needle et al., 1986) and alcohol use in specific (D'Amico and Fromme, 1997, Bahr et al., 2005, Boyle et al., 2001). All in all, previous research suggests that parents', siblings', and friends' alcohol use may be significantly related to adolescent drinking.

Most of the studies in this area have focused either on parents, siblings, or best friends, but only few have simultaneously examined the relative impact of drinking of each of these persons for adolescent drinking in one study. Comparison of parents' and peers' drinking has generally lead to the conclusion that the alcohol use of peers seems to be more important for adolescent alcohol use than parents' alcohol use. Jackson (1997) revealed that although parental alcohol use was significantly related to alcohol initiation and experimentation among adolescents, the association was much weaker than that of friends' use and number of using friends. Other studies provided further evidence for the dominance of peers' drinking over parents' drinking in that adolescent alcohol use was significantly higher related to friends' use than to parental use (Björkqvist et al., 2004, Reifman et al., 1998).

One potential limitation of some studies on parental alcohol use is that they have combined fathers' and mothers' alcohol use in one overall measure of parental alcohol use. Although the use of this measure can be informative, it may only tell part of the story since fathers and mothers may each exert a unique influence on their offspring' drinking. A small number of studies reported sex differences in the effects of parental drinking, but the findings were mixed with respect to the dominance of the effects of one parent over the other. Zhang, Welte, and Wieczorek (1999) reported that fathers' but not mothers' drinking was directly related to adolescent drinking. Seljamo et al. (2006) found that both mothers' and fathers' self-reported alcohol use was related to adolescents' problematic alcohol use, but that the fathers' alcohol use was the most potent predictor. In contrast, in the studies by Rittenhous and Miller (1984) and Marsden et al. (2005), mothers' but not fathers' drinking was positively related to adolescent drinking. Furthermore, Hundleby and Mercer (1987), Reifman et al. (1998), and Björkqvist et al. (2004) did not find any differences between the direct associations of fathers' and mothers' drinking, and adolescent alcohol use.

The relative importance of siblings' drinking compared to friends' or parents' drinking has received very little scientific attention until now. The few studies that exist on alcohol use point out that sibling effects may be similar to peer effects, and stronger than parental effects. Windle (2000) reported that when parental, sibling, and peer alcohol use were considered simultaneously in one model, the effects of the parents were weakest. Similar findings were reported by Ary et al. (1993). Moreover, in a study on parents and siblings, Boyle et al. (2001) reported small and non-significant associations between adolescents' and parental alcohol use, while the associations with sibling alcohol use were strong.

Studies that explored the associations between parental, friends' and sibling use, and adolescent alcohol use have occasionally documented on the lack of sex differences in these associations. Björkqvist et al. (2004) and Andrews et al. (2002) reported that friends' use was related to adolescents' or young adults' alcohol use, and that this association did not differ for males and females. In addition, the studies by Björkqvist et al. (2004) as well as Seljamo et al. (2006) revealed that the associations with parental alcohol use were similar for boys and girls. Thus, it seems that female and male drinking is equally related to the drinking behavior of parents, friends, and siblings.

Whereas much research has focused on early, middle, or late adolescence, relatively little is known to what extent alcohol use of fathers, mothers, siblings, and friends is related to alcohol use in young adulthood. Findings from studies on adolescents can provide some insights, but may not be entirely generalized to young adults. While parents are still influential in the lives of adolescents, in young adulthood the influence of parents often declines, while the influence of friends remains (see Bot, Engels, Knibbe, & Meeus, resubmitted). Only a very small number of studies exist that have investigated the roles of parents, siblings, or friends in relation to young adults' drinking. They reported that friends' drinking was still associated with young adults' drinking in young adulthood (Andrews et al., 2002, Labouvie, 1996, Thombs et al., 1997), while parents' drinking was not (Thombs et al., 1997). These findings reflect that friends' drinking remains important in relation to adolescents' drinking from adolescence into young adulthood, while at the same time the impact of parents' alcohol use seems to decrease in importance.

The present study is one of the first to explore the relative risks of adolescent and young adult alcohol use, directly comparing the associations between adolescent alcohol use and fathers', mothers', siblings' and friends' alcohol use. We used a genetic informative sample of monozygotic (MZ) and dizygotic (DZ) twins, allowing to substantiate possible genetic influences on relative risk for adolescent alcohol use. Based on previous research we expected to find that both fathers' and mothers' drinking would show weaker associations with adolescent drinking than siblings' and friends' drinking would. Given the mixed findings in previous studies, however, we did not anticipate a consistent pattern of sex differences in the associations of fathers' versus mothers' drinking.

Friends' and siblings' drinking were expected to show similar associations with adolescent alcohol use. In addition, we expected that the difference in risk posed by MZ and DZ same-sex twins would increase with increasing age. This hypothesis was based on behavior genetic research showing that genetic factors explain a large amount of variance in regular drinking, especially in older adolescents and young adults (Heath et al., 1991, Hopfer et al., 2003, Maes et al., 1999, Viken et al., 1999).

Since evidence indicates no sex difference in the associations between adolescents' and parents', friends, and siblings' use (e.g., Andrews et al., 2002, Björkqvist et al., 2004, Seljamo et al., 2006) we did not expect different findings for male and female adolescents. We did expect age differences in the associations, though. Based on previous research (Andrews et al., 2002, Labouvie, 1996, Thombs et al., 1997) we expected that with increasing age, the associations of adolescents' alcohol use with fathers' and mothers' use would weaken, but that the associations with friends' use would remain strong.