Developmental Patterns of Substance Use by Gender and Their Relation to Individual,Parental,and Peer Factors

Abstract

The current study aims to explore developmental pathways of substance use by gender. The influences of various predictors of gendered trajectories of substance use are also assessed. We draw on data from five waves (age 14-18 years old) of Korean Youth Panel Survey (KYPS). Semiparametric group-based models are estimated to identify heterogeneity in substance use for males and females. Multinomial logistic regression analyses are conducted to examine the association between substance use and predictors. Four alcohol use and tobacco use trajectory groups were identified for males and females. A notable difference between genders is that, unlike females, male substance use trajectories do not show a decreasing pattern. Regarding predictors of trajectory group membership, results indicate mixed findings between males and females.

Keywords

substance use developmental trajectories gender risk factors protective factors

Introduction

Excessive substance use has an adverse impact on public health. However, the detrimental effects of substance use on health are greater for adolescents than for adults because adolescence is a critical period of physical and psychological development for the individuals (Arseneault et al., 2002; Grant & Dawson, 1998; Newcomb & Bentler, 1987, 1988; Schramm-Sapyta, Walker, Caster, Levin, & Kuhn, 2009). This means that substance use and abuse have the potential to influence one’s life course.

Substance use during adolescence is a strong predictor of increased risk of substance use disorders (e.g., alcohol dependence) and serious drug use in adulthood (DeWit, Adlaf, Offord, & Ogborne, 2000; Ellickson, Hays, & Bell, 1992; Grant & Dawson, 1998; Hawkins et al., 1997; Kandel, Yamaguchi, & Chen, 1992). Numerous studies have reported that substance use is significantly related to negative consequences such as poor health (e.g., respiratory and cardiovascular diseases), mental disorders, drug use disorders, suicidal ideation, social/economic disadvantage, poor academic achievement, unemployment, disrupted familial relationships, increased accidents/injury, increased mortality, greater risk of victimization, and increased criminal behavior (Berg et al., 2013; Dawson, Li, & Grant, 2008; S. C. Duncan, Alpert, Duncan, & Hops, 1997; Exum, 2006; Hill, White, Chung, Hawkins, & Catalano, 2000; Jennison, 2004; Marshal, 2003; Mullahy & Sindelar, 1996; Özbay, 2008; Plant, Miller, Thornton, Plant, & Bloomfield, 2000; Roerecke & Rehm, 2013; Tucker, Ellickson, Orlando, Martino, & Klein, 2005; Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994; Yu, 1998; Zhang, Wieczorek, Welte, Colder, & Nochajski, 2010).

It is well documented that males are more likely to use substances compared with females across the life course. This is true in terms of both frequency and quantity of use, except during early adolescence (Buu, Dabrowska, Heinze, Hsieh, & Zimmerman, 2015; Chen & Jacobson, 2012; Johnston, O’Malley, Bachman, & Schulenberg, 2013; Muthén & Muthén, 2000). However, substance abuse in females tends to be associated with harsher outcomes compared with substance abuse in males. For example, females who abuse substances face an increased risk of comorbid psychiatric disorders (e.g., major depression, panic disorder, anxiety disorder, phobia, and posttraumatic stress disorder), and are more likely to report emotional, physical, and sexual abuse than do males (Brady & Randall, 1999; Hesselbrock, Meyer, & Keener, 1985; Kessler et al., 1997; Messina, Grella, Burdon, & Prendergast, 2007). Similarly, female offenders are more likely to repeat substance use and end up in a cycle of addictive behavior (Anthony, Warner, & Kessler, 1994; Bennett, Holloway, & Farrington, 2008; Kippin et al., 2005). In contrast with males, the negative consequences of substance use emerge earlier in females, and tend to be more severe (Nolen-Hoeksema, 2004).

Previous studies examining predictors of substance use show inconclusive results by gender (Barfield-Cottledge, 2015; Farrell & White, 1998; Flannery, Vazsonyi, Torquati, & Fridrich, 1994; Johnson & Marcos, 1988; Kilpatrick et al., 2000; Newcomb, Maddahian, & Bentler, 1986; Nolen-Hoeksema, 2004). For instance, Nolen-Hoeksema (2004) found an overall similarity between predictors of substance use for males and females. Flannery et al. (1994), alternatively, found differential effects of parental and school-related predictors, aggression, and depression on substance use patterns for Caucasian men and women. These inconsistent results highlight the need to investigate the role of gender in predicting substance use. Therefore, the current study first aims to explore whether gendered developmental trajectories of substance use exist. The second aim is to examine how individual, parental, and peer factors influence substance use for males and females. We accomplish these aims using data from a South Korean youth panel sample.

Developmental Trajectories of Substance Use by Gender

Gender issues between male and female substance use vary at different points in the life course. In early adolescence, girls show higher or similar levels of substance use than boys; however, this pattern changes throughout young adulthood (Johnston, O’Malley, Bachman, & Schulenberg, 2011; Moffitt, Caspi, Rutter, & Silva, 2001; The Substance Abuse and Mental Health Services Administration [SAMHSA], 2012). Males show greater increases in substance use over time, exceeding substance use by females during middle and late adolescence and early adulthood (Biehl, Natsuaki, & Ge, 2007; Buu et al., 2015; Chen & Jacobson, 2012; S. C. Duncan, Duncan, & Strycker, 2006; Evans-Polce, Vasilenko, & Lanza, 2015; Johnston et al., 2013; SAMHSA, 2012).

A longitudinal assessment is necessary to examine these gendered dynamics of substance use over time. Using longitudinal multilevel modeling, Chen and Jacobson (2012) explored long-term gender differences in substance use. They found a significantly higher level of substance use for women than for men at age 12. However, the prevalence of substance use for males sharply increases after early adolescence, and males’ substance use surpasses that of females by age 15. Levels of substance use are also higher in men than in women during adulthood. The gender gap for tobacco use is smaller than that for alcohol use, given that women are more likely to use tobacco than alcohol. Although the substance use of both men and women reaches its peak around the mid-20s, the stationary point for men is slightly later than that for women. These gendered patterns were consistently observed regardless of the type of substance. Evans-Polce and colleagues (2015) analyzed the National Longitudinal Study of Adolescent Health (Add Health) using a time-varying effect model to investigate gender disparities in substance use behaviors. In terms of prevalence, men and women reported similar substance use rates around age 14 to 15, whereas men exhibited higher levels of cigarette use, regular heavy episodic drinking, and marijuana use from age 16 to 32. Tobacco and marijuana use among men peaked later than for women, while heavy episodic drinking among women peaked later than among men. The gender gap for tobacco use was smaller than that for regular heavy episodic drinking.

Although these studies provide general developmental patterns of substance use by gender, they do not examine whether subgroups with different pathways may exist for both genders. Previous studies attempted to identify clusters with homogeneous features in substance use; however, studies which specifically examine heterogeneity in substance use by gender are scarce. Nelson, Van Ryzin, and Dishion (2015) relied on longitudinal data from the ages of 12 to 24, and explored developmental pathways of alcohol, marijuana, and tobacco use using a group-based trajectory model. Although this study did not specifically focus on gender differences in substance use, they identified subgroups of users by different types of substances. They identified eight trajectories in alcohol use, six trajectories in tobacco use, and seven trajectories in marijuana use. In general, being male predicted membership in the increasing use groups for alcohol, marijuana, and tobacco compared with the abstainer trajectory. Using four waves of data from ages 15 to 28, White, Johnson, and Buyske (2000) examined the effects of parental factors on the alcohol and cigarette use of children. They found four drinking trajectories (low, late, moderate, and heavy usage) in both males and females, as well as three smoking trajectories (low, moderate, and heavy usage) in both males and females. Results show similar trajectories for alcohol and tobacco use, regardless of gender. However, the gender gap for tobacco use is smaller than that for alcohol use in terms of quantity and frequency. Windle, Mun, and Windle (2005) found four alcohol trajectories (nonuse, moderate, high, and very high group) for men and five trajectories (nonuse, infrequent, time-limited, moderate, and high group) for women from ages 16 to 25. On average, men showed greater alcohol use and greater variability in frequency of alcohol use compared with women.

Although a handful of studies (e.g., Chen & Jacobson, 2012; Evans-Polce et al., 2015) provided general developmental patterns of substance use by gender, little information exists on potential subgroups which may show distinct developmental pathways and differences between the subgroups. Furthermore, there is very limited research on gendered variations in substance use trajectories. Identifying heterogeneity between substance use subgroups may help us understand the developmental course of substance use, and elucidate the correlates and comorbid behaviors of substance use for males and females.

Gender Issues in Predictors of Substance Use

Identifying risk and protective factors for substance use serves as a good starting point for explaining the variations among different developmental pathways and changes over time. Based on the previous literature, the current study includes three domains of risk and protective factors: individual, parental, and peer factors.

Individual Factors

The relationship between self-control and deviant behaviors is well documented (Gottfredson & Hirschi, 1990; Grasmick, Tittle, Bursik, & Arneklev, 1993; Pratt & Cullen, 2000). Individuals with low self-control are more likely to engage in risky behaviors. Characteristics of low self-control—such as impulsivity, sensation seeking, and behavior undercontrol—are significantly associated with substance use (Caspi et al., 1997; Cloninger, Sigvardsson, & Bohman, 1988; Sher, Trull, Bartholow, & Vieth, 1999; Waldeck & Miller, 1997; Zuckerman & Kuhlman, 2000). Although males tend to have lower levels of self-control than females, how this relates to substance use is still unclear. Caspi, Moffitt, Newman, and Silva (1996) reported a significant relationship between boys with behavioral control problems and later substance use, but no such relationship was found for girls. This gender difference was also found by Rutledge and Sher (2001). They reported that behavior undercontrol is more strongly associated with heavy drinking by men than by women. However, Slutske et al. (2002) showed no gendered effects of self-control on alcohol use.

Victimization is significantly related to substance use. Exposure to violence is a predictor of substance use; however, the strength of this association varies by gender. Kaufman (2009) reported a significant link between regular alcohol consumption and victimization of boys and girls. Kilpatrick et al. (2000) did not find gender differences in the relationship between vicarious violence and alcohol or drug use. Similarly, Pinchevsky, Wright, and Fagan (2013) compared the impacts of direct and indirect exposure to violence on substance use by teenagers. They found a significant association between violent victimization and substance use but reported no gender differences in the effects. Whaley, Hayes-Smith, and Hayes-Smith (2013) examined the impact of victimization as a mediator of substance abuse. The relationship between gender and alcohol and marijuana use was mediated by violent victimization. The association between gender and casual and heavy drinking was substantially lessened when violent victimization was included in the analysis.

Parental Factors

Family environment has been shown to be a strong predictor of substance use (Dishion, Patterson, & Reid, 1988; T. E. Duncan, Duncan, & Hops, 1994; Hops, Tildesley, Lichtenstein, Ary, & Sherman, 1990). Family cohesion has received support as a protective factor in the initiation of substance use (Hops et al., 1990; Jessor & Jessor, 1977). Fagan, Van Horn, Hawkins, and Jaki (2013) found a significant relationship between poor parental controls and the increased risk of adolescent substance use and differential influences on the instrumental parental controls (poor family management and parental approval of drug use and delinquency) toward substance use. However, there were no gender differences in the differential effects of parental controls on substance use.

The deterring effects of parental attachment on adolescent substance use have received moderate support (Bahr, Marcos, & Maughan, 1995; Baumrind, 1985; Brook, Whiteman, & Gordon, 1982; Brook, Whiteman, Gordon, & Brook, 1984; Fagan et al., 2013; Farrell & White, 1998; Hawkins, Catalano, & Miller, 1992). Overall, the impact of parental attachment on substance use appears to be stronger for girls than for boys; however, these findings are not consistent (Bahr et al., 1995; Barfield-Cottledge, 2015; Farrell & White, 1998; Johnson & Marcos, 1988).

Another important aspect of the role of parents is parental supervision or parental knowledge of youth whereabouts and activities, which has been shown to be an important protective factor against youth substance use (Deković, 1999; Dishion, Capaldi, Spracklen, & Li, 1995; Forehand, Miller, Dutra, & Chance, 1997; Kerr & Stattin, 2000; Li, Feigelman, & Stanton, 2000; Stattin & Kerr, 2000; Steinberg, Fletcher, & Darling, 1994; Westling, Andrews, Hampson, & Peterson, 2008). Parental supervision and parental knowledge act as a buffer against youth substance use by limiting the chances of engaging in risky situations and deviant peer association (Li et al., 2000). Similarly, poor parental supervision and knowledge of youth activities are significantly related to the increased risk of substance use (Dishion & Loeber, 1985; Paternoster & Triplett, 1988; Steinberg et al., 1994). Using a Swedish youth sample, Svensson (2003) reported that females received more parental supervision than males, whereas there were no gender differences in the buffering effects of parental supervision on drug use. However, the gendered effects of parental supervision on substance use are not clear, as studies show inconsistent results. Some studies have reported similar influences of parental supervision on both genders (Martens, 1997; Steinberg et al., 1994; Westling et al., 2008), but other research has shown differential impacts on men and women (Cernkovich & Giordano, 1987; Flannery et al., 1994; Griffin, Botvin, Scheier, Diaz, & Miller, 2000; Smith & Paternoster, 1987; Weintraub & Gold, 1991).

In addition, child maltreatment perpetrated by parents increases the likelihood of substance use in adolescence (Jacob & Johnson, 1997; Moran, Vuchinich, & Hall, 2004; Rodgers et al., 2004; White et al., 2000). Moran et al. (2004) found that four types of abuse (emotional, physical, sexual, and the combination of sexual and physical abuse) were significantly related to increased risk of alcohol, tobacco, and illicit drug use. Although the results show that men who experience sexual and physical abuse are more likely to engage in illicit drug use than women, there was no gender difference in the relationship between parental abuse and alcohol and tobacco use.

Peer Factors

As youth age, they attempt to gain autonomy from their parents. Although parents are still important to adolescents during this transition, the impact of parental involvement decreases, and the influence of peers becomes more powerful in shaping their behavior (Akers, 1998; Dishion, Nelson, & Bullock, 2004; Steinberg & Silverberg, 1986; Thornberry, 1987; Thornberry & Krohn, 1997; Warr, 2002). Youth associated with prosocial peers are less likely to engage in delinquent behaviors, including substance use (Akers & Cochran, 1985; Akers, Krohn, Lanza-Kaduce, & Radosevich, 1979; Pollard, Hawkins, & Arthur, 1999; Winfree & Griffiths, 1983).

Peers have been identified as one of the most significant and consistent factors in adolescent substance use (Ary, Tildesley, Hops, & Andrews, 1993; Brook, Brook, Gordon, Whiteman, & Cohen, 1990; Dupre, Miller, Gold, & Rospenda, 1995; Elliott, Huizinga, & Ageton, 1985; Hawkins et al., 1992). Peer rejection and aggression have been shown to increase the risk of substance use (Brook et al., 1982; Hops et al., 1990; Jessor & Jessor, 1977; Loeber & Dishion, 1983). Peer pressure, which includes both specific and direct coercion, is a very important predictor of adolescent substance use (Dupre et al., 1995; Farrell & White, 1998). Peers can facilitate substance use by introducing drugs, providing drugs, modeling drug-using behaviors, and shaping attitudes about drugs (Getting & Beauvais, 1986). In general, literature on gendered effects of peers on substance use is inconclusive, but some studies have reported that peers have a greater influence on males compared with females (Hops, Davis, & Lewin, 1999; Svensson, 2003). Hops and colleagues (1999) assessed the effects of peer, family, and school factors on the initiation of alcohol use. The results showed that peer factors significantly predicted alcohol use for men but not for women.

Having a friend who uses substances is regarded as the strongest predictor of adolescent substance use (Brook, Gordon, Brook, & Brook, 1989; Coombs, Paulson, & Richardson, 1991; Flannery et al., 1994; Jessor & Jessor, 1977; Wills, Windle, & Cleary, 1998). Flannery et al. (1994) examined the impacts of peers, parents, and school factors on substance use by boys and girls. The effects of parental supervision, grade point average (GPA), and school adjustment were found to vary by ethnicity. Gender differences emerged from these three predictors for Caucasians but not for Hispanics. However, peer substance use was the best predictor of substance use, and its influence was the same across gender and ethnicity.

Although many studies examined predictors of substance use, they showed mixed evidence regarding gender differences in the strength of the relationships between these risk and protective factors and substance use. A handful of studies have explored different types of predictors using group-based trajectory models. White et al. (2000) found that mothers’ drinking habits significantly influenced boys’ and girls’ drinking, whereas only sons were affected by fathers’ alcohol use. However, parental smoking habits did not predict smoking habits of their children. Windle and colleagues (2005) found that lower religious commitment and higher delinquent behaviors were common risk factors for both men and women. However, lower GPA, lower task orientation, more stressful life events, and earlier onset of drinking relatives predicted high alcohol use for males compared with that for females. However, the underlying mechanisms leading to the different trajectories are not fully described. Even fewer studies have investigated gendered effects of predictors on developmental pathways of substance use. Longitudinal investigations of whether the same predictors affect substance use for males and females further our current knowledge about gender differences in substance use. These efforts may also be useful for developing effective intervention programs.

Substance Use in South Korea

Adolescent substance use has emerged as a major social problem in South Korea. Although the legal drinking and smoking age are set at 19 years in South Korea, some youth begin using substances earlier. According to the Korean Ministry of Gender Equality and Family (2012), 47% of juveniles reported alcohol use and 24.6% of juveniles reported tobacco use. In 2007, adolescent substance use increased by 38.1% compared with the previous year (Korean Association Against Drug Abuse, 2008). The prevalence of alcohol and cigarette use among Korean youth is higher than that of American youth (Korean National Youth Policy Institute, 2010). In 2002, heavy episodic drinking rates among youth were 11.61% in Korea and 10.7% in the United States (Korean Ministry for Health Welfare and Family Affairs & Korean Association of Smoking and Health, 2008).

Despite the high prevalence of substance use among Korean adolescents, little academic research on the topic has been conducted in the South Korean context. Some Korean studies on adolescent substance use compare the relative power of parental and peer influences on substance use. Hwang and Akers (2003, 2006) reported that peers have a stronger impact on Korean adolescent substance use than parents. Relying on three waves of panel data, Park (2007) examined factors predicting cigarette use frequency. Although relationship with parents was a significant predictor of increases in frequency of smoking, risk factors for cigarette smoking included having friends who smoke and the adolescent’s own delinquent behaviors. Peer substance use is a strong predictor of substance use in both cross-sectional (Hwang, 2012) and longitudinal studies (Hwang, 2010), whereas parental factors did not exert a significant influence in either case.

On the contrary, a longitudinal study by E. Kim, Kwak, and Yun (2010) showed that parental and peer factors exert similar influences on substance use among Korean adolescents. These somewhat inconsistent results may reflect the cultural duality present in South Korean society. Contemporary Korean society is influenced by traditional Korean culture and its emphasis on collectivism, Confucianism, and parental authority as well as being influenced by Western culture. The current parental generation was raised and educated in traditional ways, whereas younger generations are more exposed to Western culture and values. Despite the attempts of adolescents to reject parental influence, Korean parents typically attempt to pass on traditional Korean values to their children. Therefore, traditional values pursued by parents inside the home contrast with Western culture outside the home. This cultural dualism may lead to a more equal influence of parents and peers on youth behaviors.

The influence of gender on substance use has not been fully explored in Korean studies, with only a few studies addressing the link between gender and substance use. W. Kim (2014) examined risk factors of adolescent drinking by gender, and found that stress and depression are common predictors of alcohol consumption for both adolescent men and women. The study also revealed some differences by gender. Specifically, attachment to family predicted alcohol use for males but not for females. Using longitudinal data, Chun and Chung (2013) found that poor parental attachment was only a significant predictor of male smoking habits, whereas peer substance use and low self-control were common risk factors for both male and female substance use. Cho and Yoon (2010) used a latent growth curve model, and investigated the relationship between parental supervision and adolescent substance use. They found a bidirectional relationship between supervision and substance use; however, they did not observe gender differences in these outcomes. Y. Kim, Kim, and Kim (2001) found gender differences in the link between peer substance use and one’s own substance use, whereas parental supervision exerted a significant impact only for girls.

The majority of Korean studies on substance use have relied on cross-sectional data, which cannot reveal developmental patterns that may appear over time. It is, therefore, imperative to conduct longitudinal research on patterns of substance use in South Korea. In addition, the influences of gender on substance use and the identification of heterogeneous subgroups have received very limited attention in South Korea.

The Current Study

Despite academic efforts to understand gendered effects on adolescent substance use, the link between gender and substance use remains unclear. In addition, the majority of studies which address gender and substance use draw on American samples, paying little attention to different cultural contexts. To address this gap in the literature, this study examines the relationship between gender and substance use by analyzing five waves of data on Korean youth. The purpose of this study was to investigate various risk and protective factors of gendered substance use from life-course perspective. To the best of our knowledge, this study is the first effort to investigate latent groups of male and female substance users and gendered predictors of substance use utilizing a Korean sample.

Method

Data

The current study draws on the Korean Youth Panel Survey (KYPS). The KYPS is a study conducted by the National Youth Policy Institution (NYPI) to provide longitudinal data in South Korea. The KYPS used stratified multistage cluster sampling to increase the representativeness of the sample. The Korean youth population was divided into three categories: (a) those living in the Seoul metropolitan area (the capital of South Korea and the country’s largest city), (b) those living in a compilation of six other metropolitan cities, and (c) those living in other cities and counties equally drawn from all nine Korean provinces. The KYPS calculated population proportions of the three categories among the entire Korean population and the sample drew from representative schools. In all, 104 schools were randomly selected from each cluster, and one class was randomly selected from each school. The KYPS surveyed students in the second year of junior high school every year from 2003 (14 years old) to 2008 (19 years old).

Data collection proceeded in three steps: First, students gave paper-and-pencil responses to questions about problematic behaviors. Second, well-trained interviewers interviewed respondents at school to gain more information about adolescents and their behaviors. Third, the respondents’ parents participated in a telephone survey to gather information on family background, the parents’ education level, and the family’s socioeconomic status.

Although the KYPS consisted of six waves, this study analyzes only Waves 1 through 5. Analysis does not include Wave 6, because the contents of questionnaire were largely changed, and some of measures we employed were not available. There was some attrition during Wave 1, but most participants who completed the first- and second-year surveys remained in the study. After excluding attrition cases, approximately 80% of the original sample was left for this study. There were no significant differences between the cases analyzed in the current study and attrition cases. The number of respondents was 2,721, and the ratio of gender was almost equal (1,370 men and 1,351 women). There was no diversity in race or ethnicity, as a characteristic of South Korea is ethnic homogeneity.

Measures

Dependent Variables

Alcohol use

Alcohol use was measured across all five waves. Respondents were asked how frequently they had consumed alcohol during past 30 days (J. Lee, 2012). Response options were never (0), less than once per month (1), less than once per 2 weeks (2), less than once per week (3), 2 or 3 times per week (4), and daily use (5).

Tobacco use

Tobacco use was measured across all five waves. Respondents were asked how frequently they smoked during past 30 days (J. Lee, 2012). Response options were never (0), less than once per month (1), less than once per 2 weeks (2), less than once per week (3), 2 or 3 times per week (4), and daily use (5).

Individual Factors

Self-control

Self-control was measured with six items in Wave 1. Respondents answered the following questions on a 5-point Likert-type scale: (a) I easily abandon a difficult task; (b) I don’t do my homework habitually; (c) Even if I have an exam tomorrow, I jump into exciting things; (d) I am apt to enjoy risky activities; (e) I enjoy teasing and harassing other people; and (f) I lose my temper whenever I get angry (Y. Lee & Kim, 2016). These questions represent dimensions of low self-control, including impulsivity, abandoning simple tasks, risk taking, self-centeredness, and having a temper. The internal consistency of low self-control was .620 for male and .681 for female adolescents. High scores indicate low self-control.

Victimization

In Wave 1, participants reported whether they had been victimized in five different ways: (a) threatened in the last year, (b) bullied in the last year, (c) seriously beaten up in the last year, (d) sexually assaulted or harassed in the last year, and (e) teased or ridiculed in the last year (Y. Lee & Kim, 2016). Each type of victimization was coded as a binary question (no = 0, yes = 1). A victimization variable was created with the sum of the five questions, ranging from 0 to 5.

Parental Factors

Parental attachment

Parental attachment was measured with two items on a 5-point Likert-type scale (Wave 1): (a) I am comfortable sharing my thoughts and feelings with my parents, and (b) I often talk with my parents about what happens to me outside home (Y. Lee & Kim, 2016). Cronbach’s alpha value was .694 for men and .761 for women. Higher values indicate greater parental attachment.

Parental knowledge

Parental knowledge about adolescents’ whereabouts and activities was composed of four items measured on a 5-point Likert-type scale (Wave 1): (a) When I go out, my parents usually know where I am; (b) When I go out, my parents usually know whom I am with; (c) When I go out, my parents usually know what I am doing; (d) When I go out, my parents usually know when I will return (Han, Grogan-Kaylor, 2013; Y. Lee & Kim, 2016). Cronbach’s alpha values were .823 and .873 for each men and women, respectively. Greater scores indicate more parental knowledge.

Parental abuse

Parental abuse was composed of two items measured with a 5-point Likert-type scale (Wave 1): (a) I often receive verbal abuse from my parents and (b) I am often badly hit by my parents (Han & Grogan-Kaylor, 2013). Internal consistency as measured with Cronbach’s alpha was .746 for men and .764 for women. Higher values indicate greater parental abuse.

Peer Factors

Peer attachment

Four items were used to measure peer attachment on a 5-point Likert-type scale (Wave 1): (a) I want to maintain the friendships I have with my close friends, (b) I am happy with my friends, (c) I try to have same thoughts and feelings as my friends, and (d) I have candid conversations with my friends (Y. Lee & Kim, 2016). Cronbach’s alpha value was .772 for men and .748 for women. Greater scores indicate more peer attachment.

Peer substance use

Peer substance use was measured by peer alcohol use and peer tobacco use. In Wave 1, participants were asked how many of their peers consumed alcohol and smoked in the last year, respectively.

Control Factors

Three control variables that may account for confounding effects were included in the analyses: paternal education level, maternal education level, and family income. These three variables were measured in Wave 1. Education level was divided into eight categories: no schooling, elementary school, middle school, high school, junior college, university degree, master’s degree, and PhD degree. Family income was measured in Korean currency (roughly converted as 1-to-10 U.S. dollars). Monthly family income was US$2,942 for men and US$2,858 for women.

Analytic Strategy

Analysis proceeded in four stages: First, we assigned descriptive statistics and means-differences tests by gender to attain general description of the dataset and gender differences. Second, we examined four semiparametric group-based models by separating alcohol and tobacco use for each gender. We analyzed Waves 1 through 5 with six different cubic specifications to identify the number of latent trajectories by using “traj” commands in STATA 13 (Jones & Nagin, 2012). To determine the optimal number of trajectory groups, we assessed the different orders of polynomials and Bayesian information criterion (BIC). As the value of BIC approaches 0, the model shows better model fit (Nagin, 2005). Next, we estimated the mean posterior probability to assess the fitness of the selected models (Nagin, 2005). Third, we examined the impacts of individual, parental, and peer factors on alcohol and tobacco use trajectories using multinomial logistic regression. We computed bivariate relative risk ratios (RRRs) between individual, parental, and peer factors among groups. After the series of multinomial logistic regression models, binned residual plot was examined to evaluate the model fit. In addition, multivariate outliers were checked by Studentized residuals to avoid any misleading results. Results from diagnostic statistics indicate that all values were within an acceptable range. For instance, binned residuals were placed within 95% error bounds, and values of Studentized residuals were within ranges of −2 to 2. Fourth, we conducted a series of ANOVAs and Tukey’s b tests to identify the covariates that possibly explain distinguishable decreasing groups among women. We used multiple imputation to handle missing values in the analysis.

Results

Gendered descriptive statistics and means-differences tests are reported in Table 1. The gendered differences found in most of the variables support the importance of gendered analyses of substance use. Females were more likely to use alcohol during Waves 1 and 2, but the reverse was true after Wave 3, with males reporting more use. Tobacco use indicated no gendered differences during Waves 1 and 2, but by Wave 3, tobacco use became more prevalent among males than among females. In addition, most of the bivariate correlations were statistically significant (see more details in the Appendix).

Table 1:

Descriptive Statistics and Means-Differences Test by Gender

		Female (N = 1,351)				Male (N = 1,370)
Variable	Wave	M	SD	Minimum	Maximum	M	SD	Minimum	Maximum
Alcohol use***	1	0.400	0.559	0	4	0.231	0.463	0	4
Alcohol use***	2	0.350	0.552	0	4	0.279	0.552	0	4
Alcohol use	3	0.410	0.622	0	5	0.448	0.742	0	5
Alcohol use***	4	0.509	0.746	0	5	0.685	0.963	0	5
Alcohol use***	5	0.569	0.765	0	5	0.880	1.046	0	5
Tobacco use	1	0.246	0.870	0	5	0.263	0.923	0	5
Tobacco use	2	0.270	1.046	0	5	0.340	1.170	0	5
Tobacco use***	3	0.248	1.016	0	5	0.505	1.430	0	5
Tobacco use***	4	0.287	1.116	0	5	0.756	1.736	0	5
Tobacco use***	5	0.300	1.155	0	5	0.926	1.893	0	5
Individual factors
Low self-control*	1	9.900	4.000	0	24	10.224	3.909	0	24
Victimization***	1	0.261	0.665	0	6	0.412	0.840	0	6
Parental factors
Parental attachment***	1	4.618	1.970	0	8	3.989	1.874	0	8
Parental knowledge***	1	9.352	3.414	0	16	8.459	3.378	0	16
Parental abuse*	1	1.325	1.724	0	8	1.475	1.741	0	8
Peer factors
Peer attachment***	1	12.554	2.403	0	16	11.906	2.716	0	16
Peer alcohol use***	1	1.353	2.655	0	20	0.948	3.037	0	34
Peer tobacco use*	1	0.545	1.652	0	18	0.771	2.901	0	56
Control factors
Paternal education level	1	3.654	1.270	0	7	3.700	1.294	0	7
Maternal education level	1	3.159	1.038	0	7	3.201	1.092	0	7
Family income	1	285.84	183.391	5	2,000	294.237	197.082	30	3,000

p < .05. **p < .01. ***p < .001.

Males exhibited lower level of self-control on average, experienced more victimization, and were more likely to be exposed to parental abuse than females. Females had a stronger attachment to their peers and parents, as well as higher levels of parental knowledge. Females were more likely to associate with peers who use alcohol, whereas males were more likely to associate with peers who use tobacco at Wave 1. Females indicated a stronger peer association than males. Family income and parental education levels were similar between both male and female groups.

Figure 1 illustrates alcohol use trajectories for females. The best fitting model of female alcohol use trajectory groups was determined by mutual considerations of BIC scores and posterior probabilities. The best fitting model has four latent trajectory groups (BIC = −6,469.01), and posterior probabilities are higher than the .70 threshold suggested by Nagin (2005). Four groups indicate generally similar proportions ranging between 19.10% and 29.02%. The biggest group is the midstable alcohol use group (29.02%) that reported drinking less than once per month. The mid-decreasing group accounts for 19.1% of the total sample, comprising the smallest trajectory group.

Figure 1:

Group-Based Trajectories of Female Alcohol Use

Figure 2 displays tobacco use trajectories for females. The best fitting model of female alcohol use trajectory groups was determined by mutual considerations of BIC scores and posterior probabilities. The chosen best fitting model had four latent trajectory groups (BIC = −2,301.61), and all posterior probabilities were more than the .70 threshold. Similar to the alcohol trajectory model, the female tobacco use trajectory model indicated four latent groups, but differences were found for the ratios of each trajectory group membership. The largest group is the nonused group, which accounts for 89.56% of the sample. The remaining 10% of females exhibit the following different trajectories: low-increasing (4%), mid-decreasing (4.07%), and mid-increasing (2.37%).

Figure 2:

Group-Based Trajectories of Female Tobacco Use

Figure 3 illustrates alcohol use trajectories for males. The best fitting model for male alcohol use trajectory groups was determined by the same process of mutual considerations of BIC scores and posterior probabilities. The chosen best fitting model had four latent trajectory groups (BIC = −6,730.60), and each posterior probability was more than the .70 threshold. Although the proportion of males falling into the nonalcohol use group is similar to females, two differences were found in males: (a) no decreasing groups and (b) a mid-increasing group. The biggest group is the low-increasing group (39.27%), and the smallest group is the mid-increasing group (4.89%).

Figure 3:

Group-Based Trajectories of Male Alcohol Use

Figure 4 shows tobacco use trajectories among males. The chosen best fitting model had four latent trajectory groups (BIC = −3,501.07), and posterior probabilities were more than the .70 threshold. Similar to findings from the male alcohol use trajectories, the male tobacco use trajectories also indicate no decreasing group. The largest group is the nonuse group, which accounts for 75.4% of the sample.

Figure 4:

Group-Based Trajectories of Male Tobacco Use

Tables 2 to 5 summarize the results of multinomial logistic regressions examining the effects of individual, parental, and peer factors on alcohol and tobacco use by gender. The nonused group is used as the reference group for all multinomial logistic regression models. Table 2 summarizes the results for the female alcohol use model. Low self-control (low-increasing, RRR = 1.045, SE = .020, p = .019; mid-decreasing, RRR = 1.078, SE = .023, p < .001; mid-stable, RRR = 1.109, SE = 0.022, p < .001) and peer alcohol use (mid-decreasing, RRR = 1.834, SE = 0.155, p < .001; mid-stable, RRR = 1.892, SE = 0.157, p < .001) were identified as risk factors, whereas parental knowledge (low-increasing, RRR = 0.933, SE = 0.024, p = .006; mid-decreasing, RRR = 0.895, SE = 0.025, p < .001; mid-stable, RRR = 0.883, SE = 0.024, p < .001) served as a protective factor against substance use by females.

Table 2:

Bivariate RRR Between Factors and Each Pair of Trajectory Group Comparisons: Female Alcohol

	Non vs. low-increasing			Non vs. mid-decreasing			Non vs. mid-stable
Female alcohol	RRR	SE	p value	RRR	SE	p value	RRR	SE	p value
Individual factors
Low self-control	1.045	0.020	.019*	1.078	0.023	.000***	1.109	0.022	.000***
Victimization	0.855	0.132	.310	1.091	0.163	.560	1.214	0.167	.158
Parental factors
Parental attachment	1.015	0.046	.735	0.974	0.049	.606	0.912	0.043	.050*
Parental knowledge	0.933	0.024	.006**	0.895	0.025	.000***	0.883	0.024	.000***
Parental abuse	1.061	0.053	.237	0.959	0.055	.460	1.043	0.054	.410
Peer factors
Peer attachment	1.043	0.026	.091	1.012	0.029	.683	1.022	0.027	.414
Peer alcohol use	1.150	0.111	.147	1.834	0.155	.000***	1.892	0.157	.000***
Peer tobacco use	0.794	0.130	.161	0.889	0.121	.388	0.917	0.122	.512
Control factors
Paternal education level	0.913	0.077	.280	0.846	0.082	.084	0.865	0.079	.112
Maternal education level	0.871	0.091	.186	1.034	0.122	.776	1.025	0.115	.828
Family income	1.000	0.000	.697	1.001	0.001	.300	1.000	0.001	.591

Note. F = 9.03, p < .001. RRR = relative risk ratio.

p < .05. **p < .01. ***p < .001.

Table 3:

Bivariate RRR Between Factors and Each Pair of Trajectory Group Comparisons: Female Tobacco

	Non vs. low-increasing			Non vs. mid-decreasing			Non vs. mid-increasing
Female tobacco	RRR	SE	p value	RRR	SE	p value	RRR	SE	p value
Individual factors
Low self-control	1.105	0.039	.005**	1.013	0.036	.722	1.161	0.056	.002**
Victimization	1.194	0.206	.306	1.474	0.226	.012*	1.332	0.253	.131
Parental factors
Parental attachment	0.876	0.069	.093	0.834	0.066	.023*	0.776	0.082	.016*
Parental knowledge	0.893	0.040	.013*	0.889	0.041	.011*	0.847	0.051	.006**
Parental abuse	0.892	0.076	.177	1.023	0.080	.769	1.063	0.102	.522
Peer factors
Peer attachment	0.893	0.041	.014*	0.937	0.044	.161	0.860	0.054	.017*
Peer alcohol use	1.108	0.066	.083	1.220	0.066	.000***	1.138	0.091	.105
Peer tobacco use	1.336	0.120	.001**	1.346	0.110	.000***	1.439	0.151	.001**
Control factors
Paternal education level	0.775	0.130	.128	0.939	0.157	.707	0.754	0.173	.220
Maternal education level	1.252	0.251	.263	1.004	0.206	.983	1.274	0.345	.370
Family income	0.998	0.001	.101	0.999	0.001	.246	0.996	0.002	.047*

Note. F = 26.45, p < .001. RRR = relative risk ratio.

p < .05. **p < .01. ***p < .001.

Table 4:

Bivariate RRR Between Factors and Each Pair of Trajectory Group Comparisons: Male Alcohol

	Non vs. low-increasing			Non vs. mid-stable			Non vs. mid-increasing
Male alcohol	RRR	SE	p value	RRR	SE	p value	RRR	SE	p value
Individual factors
Low self-control	1.041	0.017	.017*	1.081	0.020	.000***	1.130	0.035	.000***
Victimization	0.943	0.083	.506	1.106	0.101	.270	1.061	0.163	.699
Parental factors
Parental attachment	0.950	0.039	.213	0.873	0.040	.003**	0.881	0.071	.116
Parental knowledge	0.984	0.022	.477	0.933	0.023	.006**	0.864	0.039	.001**
Parental abuse	0.964	0.040	.377	0.959	0.044	.356	0.974	0.074	.733
Peer factors
Peer attachment	1.053	0.023	.016*	1.063	0.026	.012*	0.995	0.041	.912
Peer alcohol use	0.937	0.074	.406	1.333	0.091	.000***	1.423	0.108	.000***
Peer tobacco use	1.143	0.097	.117	1.134	0.094	.130	1.119	0.101	.216
Control factors
Paternal education level	0.921	0.069	.268	0.990	0.082	.904	0.888	0.132	.424
Maternal education level	0.924	0.083	.380	0.756	0.076	.005**	0.641	0.117	.015*
Family income	1.000	0.000	.275	1.001	0.000	.049*	1.001	0.001	.180

Note. F = 11.35, p < .001. RRR = relative risk ratio.

p < .05. **p < .01. ***p < .001.

Table 5:

Bivariate RRR Between Factors and Each Pair of Trajectory Group Comparisons: Male Tobacco

	Non vs. low-increasing			Non vs. mid-stable			Non vs. mid-increasing
Male tobacco	RRR	SE	p value	RRR	SE	p value	RRR	SE	p value
Individual factors
Low self-control	1.012	0.022	.568	1.066	0.027	.012*	1.102	0.029	.000***
Victimization	0.831	0.107	.149	1.215	0.134	.076	0.890	0.123	.397
Parental factors
Parental attachment	0.864	0.046	.006**	0.808	0.053	.001**	0.884	0.058	.063
Parental knowledge	0.946	0.027	.054	0.927	0.033	.035*	0.848	0.031	.000***
Parental abuse	0.918	0.051	.129	1.007	0.061	.905	0.980	0.060	.748
Peer factors
Peer attachment	1.002	0.029	.933	0.940	0.032	.065	0.992	0.035	.811
Peer alcohol use	0.989	0.064	.863	1.104	0.048	.025*	1.099	0.049	.033*
Peer tobacco use	1.016	0.089	.854	1.231	0.068	.000***	1.246	0.068	.000***
Control factors
Paternal education level	1.023	0.101	.821	0.815	0.102	.101	0.818	0.104	.113
Maternal education level	0.940	0.113	.606	0.954	0.143	.753	0.753	0.115	.063
Family income	0.999	0.001	.356	1.000	0.001	.478	1.000	0.001	.559

Note. F = 32.33, p < .001. RRR = relative risk ratio.

p < .05. **p < .01. ***p < .001.

Table 3 shows a more complex explanation for tobacco use patterns among females. Low self-control (low-increasing, RRR = 1.105, SE = 0.039, p = .005; mid-increasing, RRR = 1.161, SE = 0.056, p = .002) and peer tobacco use (low-increasing, RRR = 1.336, SE = 0.120, p = .001; mid-decreasing, RRR = 1.346, SE = 0.110, p < .001; mid-increasing, RRR = 1.439, SE = 0.151, p = .001) are significant risk factors, but some differences exist. Victimization (RRR = 1.474, SE = 0.226, p = .012) is strongly associated with mid-decreasing group membership when compared with the nonused group. Parental attachment (RRR = 0.776, SE = 0.082, p = .016) and peer knowledge (RRR = 0.847, SE = 0.051, p = .006) are linked to mid-increasing group participation, which indicates most the habitual consumption of tobacco. Also, peer alcohol use (RRR = 1.220, SE = 0.066, p < .001) explains female tobacco use only for the mid-decreasing group compared with nonused group.

Table 4 shows the results of multinomial logistic regression for alcohol use among males. It shows some similarities with the alcohol use among females for the explanations of low self-control (low-increasing, RRR = 1.041, SE = 0.017, p = .017; midstable, RRR = 1.081, SE = 0.020, p < .001; mid-increasing, RRR = 1.130, SE = 0.035, p < .001), peer alcohol use (midstable, RRR = 1.333, SE = 0.091, p < .001; mid-increasing, RRR = 1.423, SE = 0.108, p < .001), and parental knowledge (midstable, RRR = 0.933, SE = 0.023, p = .006; mid-increasing, RRR = 0.864, SE = 0.039, p = .001). However, there are differences in the role of parental and peer attachment. Higher parental attachment prevents consistent alcohol use by men (midstable, RRR = 0.873, SE = 0.040, p = .003), but higher peer attachment affects both males’ initiation (low-increasing, RRR = 1.053, SE = 0.023, p = .016) and a lower level of consistent use (midstable, RRR = 1.063, SE = 0.026, p = .012).

Table 5 indicates the bivariate RRR between factors and each pair of male tobacco trajectory group comparisons. General similarities are shown with the female tobacco use model for the following factors: low self-control (midstable, RRR = 1.066, SE = 0.027, p = .012; mid-increasing, RRR = 1.102, SE = 0.029, p < .001), peer tobacco use (midstable, RRR = 1.231, SE = 0.068, p < .001; mid-increasing, RRR = 1.246, SE = 0.068, p < .001), parental attachment (low-increasing, RRR = 0.864, SE = 0.046, p = .006; midstable, RRR = 0.808, SE = 0.053, p = .001), and parental knowledge (midstable, RRR = 0.927, SE = 0.033, p = .035; mid-increasing, RRR = 0.848, SE = 0.031, p < .001). However, peer attachment is no longer a significant predictor among tobacco use group distinctions for males.

In addition, we conducted a series of ANOVAs to further examine the female decreasing groups for alcohol and tobacco. We used Tukey’s b tests to make specific comparisons between groups. Table 6 compares factor changes from Waves 1 to 5 for female alcohol use. Prior multinomial regression analysis identified low self-control, parental knowledge, and peer alcohol use as predictors of mid-decreasing group membership when compared with the nonuse group. The results of Tukey’s b tests indicated distinguishable changes for parental knowledge and peer alcohol use. Female mid-decreasing alcohol use group members initially showed similar levels of parental knowledge to the midstable group, but differences became significant at Wave 2. The mid-decreasing alcohol use group’s peers also showed decreasing patterns, whereas peers associated with the midstable group displayed relatively stable alcohol consumption.

Table 6:

Mean Differences of Covariates Between Female Alcohol Trajectory Groups

Variable	Wave	G1	G2	G3	G4	F test	Tukey’s b
Individual factors
Low self-control	1	8.342	9.328	10.309	11.433	51.22***	All differences significant
	2	8.584	9.878	10.203	11.510	38.62***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
	3	8.179	9.742	9.606	11.121	42.15***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
	4	8.387	9.924	9.297	11.174	40.42***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
	5	8.014	9.559	9.614	10.691	33.43***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
Victimization	1	0.182	0.173	0.267	0.377	9.55***	G3 > G1; G4 > G2
	2	0.071	0.076	0.110	0.160	3.52*	G4 > G1; G4 > G2
	3	0.026	0.052	0.017	0.080	2.66*	None
	4	0.011	0.033	0.025	0.047	0.98	None
	5	0.011	0.015	0.004	0.033	1.42	None
Parental factors
Parental attachment	1	4.900	4.763	4.661	4.231	8.78***	G4 < G1; G4 < G2
	2	5.108	4.881	5.000	4.686	4.35**	G4 < G1
	3	5.239	4.985	5.047	4.796	4.15**	G4 < G1
	4	5.131	4.815	5.140	4.749	4.70**	G4 < G1; G4 < G3
	5	5.302	5.149	5.364	4.901	3.86**	G4 < G1; G4 < G3
Parental knowledge	1	10.296	9.562	8.936	8.601	19.62***	G2 < G1; G3 < G1; G4 < G1; G4 < G2
	2	10.766	10.164	10.081	9.427	11.17***	G4 < G1; G4 < G2; G4 < G3
	3	10.903	9.930	9.941	9.537	13.00***	G2 < G1; G3 < G1; G4 < G1
	4	10.926	10.030	10.246	9.793	9.91***	G2 < G1; G3 < G1; G4 < G1
	5	11.296	10.228	10.576	9.802	12.78***	G2 < G1; G3 < G1; G4 < G1
Parental abuse	1	0.997	1.286	1.169	1.656	10.67***	G4 > G1; G4 > G2; G4 > G3
	2	0.980	1.219	1.254	1.612	8.28***	G4 > G1; G4 > G2; G4 > G3
	3	0.772	1.085	0.983	1.361	10.62***	G2 > G1; G4 > G1; G4 > G2; G4 > G3
	4	0.875	1.097	1.153	1.427	9.30***	G4 > G1; G4 > G2
	5	0.832	1.103	1.034	1.331	7.54***	G4 > G1; G4 > G3
Peer factors
Peer attachment	1	12.274	12.441	12.547	12.884	3.97**	G4 > G1
	2	12.618	12.948	13.352	13.284	6.26***	G3 > G1; G4 > G1
	3	13.063	13.258	13.419	13.452	2.13	None
	4	12.906	13.240	13.326	13.176	1.94	None
	5	12.963	13.237	13.708	13.493	5.37**	G3 > G1
Peer alcohol use	1	0.108	0.094	0.703	1.182	36.38***	G3 > G1; G4 > G1; G3 > G2; G4 > G2; G4 > G3
	2	0.085	0.140	0.504	1.413	22.61***	G4 > G1; G4 > G2; G4 > G3
	3	0.111	0.350	0.356	1.292	44.68***	G4 > G1; G4 > G2; G4 > G3
	4	0.037	0.286	0.191	1.311	63.03***	G4 > G1; G4 > G2; G4 > G3
	5	0.020	0.316	0.394	1.391	55.30***	G3 > G1; G4 > G1; G4 > G2; G4 > G3
Peer tobacco use	1	0.262	0.353	2.102	2.818	95.80***	G3 > G1; G4 > G1; G3 > G2; G4 > G2; G4 > G3
	2	0.199	0.322	1.733	3.077	63.41***	G3 > G1; G4 > G1; G3 > G2; G4 > G2; G4 > G3
	3	0.499	1.632	1.284	3.609	82.51***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
	4	0.427	2.149	0.775	3.661	106.46***	G2 > G1; G4 > G1; G3 < G2; G4 > G2; G4 > G3
	5	0.325	2.347	1.386	4.331	128.36***	All differences significant

Note. G1 = nonused; G2 = low-increasing; G3 = mid-decreasing; G4 = mid-stable.

p < .05. **p < .01. ***p < .001.

Table 7 summarizes mean comparisons for the female tobacco use trajectory groups. Decreasing patterns of female tobacco use are explained by many factors observed in previous multinomial regression analysis: victimization, parental attachment, parental knowledge, peer alcohol use, and peer tobacco use. Victimization, peer alcohol, and tobacco use are all significant, which indicates consistent diminishing patterns. Parental attachment and parental knowledge are limited to suggest distinguishable changes for mid-decreasing groups among female tobacco use trajectories.

Table 7:

Mean Differences of Covariates Between Female Tobacco Trajectory Groups

Variable	Wave	G1	G2	G3	G4	F test	Tukey’s b
Individual factors
Low self-control	1	9.475	12.981	12.191	15.074	42.15***	G2 > G1; G3 > G1; G4 > G1; G4 > G3
	2	9.678	12.827	13.213	13.889	38.44***	G2 > G1; G3 > G1; G4 > G1
	3	9.393	11.808	12.106	13.593	27.41***	G2 > G1; G3 > G1; G4 > G1
	4	9.445	12.596	11.766	13.370	24.78***	G2 > G1; G3 > G1; G4 > G1
	5	9.133	13.000	11.851	12.778	31.22***	G2 > G1; G3 > G1; G4 > G1
Victimization	1	0.217	0.365	0.702	0.704	14.99***	G3 > G1; G4 > G1
	2	0.081	0.096	0.511	0.444	22.27***	G3 > G1; G4 > G1; G3 > G2; G4 > G2
	3	0.033	0.212	0.064	0.259	10.31***	G2 > G1; G4 > G1
	4	0.021	0.115	0.085	0.148	5.82***	G2 > G1; G4 > G1
	5	0.013	0.096	0.000	0.074	3.67*	G2 > G1; G3 > G2
Parental factors
Parental attachment	1	4.700	4.192	4.043	3.556	6.01***	G4 < G1
	2	4.959	4.404	4.340	4.778	4.35**	G3 < G1
	3	5.062	4.519	4.766	4.296	3.84**	G4 < G1
	4	4.996	4.212	4.745	4.444	4.07**	G2 < G1
	5	5.212	4.596	4.787	4.704	2.82*	None
Parental knowledge	1	9.539	8.308	7.936	6.963	11.00***	G2 < G1; G3 < G1; G4 < G1
	2	10.221	9.019	9.128	8.926	6.30***	G2 < G1; G4 < G1
	3	10.224	8.904	9.277	7.963	8.89***	G2 < G1; G4 < G1
	4	10.362	8.635	9.553	9.741	7.14***	G2 < G1
	5	10.614	8.615	9.681	9.000	8.56***	G2 < G1; G4 < G1
Parental abuse	1	1.205	1.500	2.191	2.963	11.71***	G3 > G1; G4 > G1; G4 > G2
	2	1.166	2.000	2.511	2.222	13.47***	G2 > G1; G3 > G1; G4 > G1
	3	0.972	1.615	1.915	2.185	14.68***	G2 > G1; G3 > G1; G4 > G1
	4	1.078	1.442	1.553	2.481	9.87***	G4 > G1; G4 > G2; G4 > G3
	5	0.980	1.904	1.681	2.741	23.09***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
Peer factors
Peer attachment	1	12.463	12.885	13.681	13.185	4.19**	G3 > G1
	2	12.969	13.538	13.745	13.296	2.61*	None
	3	13.265	13.673	13.489	13.259	0.71	None
	4	13.135	13.615	12.830	13.259	1.00	None
	5	13.278	14.058	13.532	13.370	2.00	None
Peer alcohol use	1	0.292	1.615	3.064	3.667	96.88***	G2 > G1; G3 > G1; G4 > G1; G3 > G2; G4 > G2
	2	0.270	1.654	4.085	4.407	70.81***	G2 > G1; G3 > G1; G4 > G1; G3 > G2; G4 > G2
	3	0.305	1.981	2.170	5.556	180.87***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
	4	0.213	2.442	2.213	5.593	252.74***	G2 > G1; G3 > G1; G4 > G1; G4 > G2; G4 > G3
	5	0.288	3.673	1.489	4.296	174.10***	All differences significant
Peer tobacco use	1	1.010	3.154	5.426	5.296	82.25***	G2 > G1; G3 > G1; G4 > G1; G3 > G2; G4 > G2
	2	0.948	3.808	6.000	4.778	56.90***	G2 > G1; G3 > G1; G4 > G1; G3 > G2
	3	1.549	3.673	4.021	5.889	40.62***	G2 > G1; G3 > G1; G4 > G1; G4 > G2
	4	1.536	4.673	4.149	5.926	55.01***	G2 > G1; G3 > G1; G4 > G1
	5	1.907	5.962	2.851	5.259	39.04***	G2 > G1; G4 > G1; G3 < G2; G4 > G3

Note. G1 = nonused; G2 = low-increasing; G3 = mid-decreasing; G4 = mid-increasing.

p < .05. **p < .01. ***p < .001.

Discussion

This study utilizes five waves of a longitudinal panel dataset to examine developmental latent patterns of alcohol and tobacco use among South Korean juveniles by gender. The present study found gendered patterns for alcohol and tobacco use. This also suggests that some factors are differently associated with developmental trajectories of alcohol and tobacco use by gender. Similarities between developmental patterns of alcohol and tobacco use within same gender groups were identified. For males, these were nonused, low-increasing, mid-stable, and mid-increasing groups for both alcohol and tobacco use. For females, they were nonused, low-increasing, mid-decreasing, and mid-increasing group trajectories.

Shapes of the trajectories were similar within each gender, but proportions of members in the groups differed. Tobacco was less popular than alcohol for both males and females. The tobacco nonused group was the largest trajectory group for females (89.56%) and males (75.4%), while alcohol nonused groups accounted for only 26.79% of females and 27.3% of males. This finding is consistent with previous studies reporting greater substance use among males than among females (Buu et al., 2015; Chen & Jacobson, 2012; Johnston et al., 2013; Muthén & Muthén, 2000; SAMHSA, 2012). However, unlike Western samples (Chen & Jacobson, 2012; Evans-Polce et al., 2015), Korean males and females are more likely to use alcohol than tobacco. This may be due to a unique drinking culture in South Korea (Chi, Lubben, & Kitano, 1989). In general, Koreans have permissive attitudes toward alcohol consumption, and alcohol-related deviance (e.g., driving under the influence [DUIs]) used to be punished much less severely. In spite of increased public awareness of the potential harms of alcohol consumption nowadays, overall Korean society still treats alcohol permissively.

The most notable difference between the substance use trajectories for males and females was the nonexistence of a decreasing group for males. Compared with the female decreasing groups, there are large increasing groups in the male sample. Previous studies that have examined latent groups of substance use by gender have partially supported this finding (White et al., 2000; Windle et al., 2005). Windle et al.’s (2005) research indicated complicated patterns of heavy drinking among females, including nonstable, infrequent-stable, moderate-stable, high, and time-limited patterns. One third of the female sample (34%) indicated a time-limited pattern that peaks between age 19 and 21, and then decreased. Similarly, White and colleagues (2000) found a similar latent group of smoking and alcohol trajectories peaking at age 20, but it was relatively stable and showed similar patterns for both genders.

More in-depth examinations using ANOVAs and Tukey’s b tests allowed us to consider time-varying changes in key factors. Results highlighted the important effects of both parental knowledge of youth whereabouts and activities and peer alcohol use on membership into the female alcohol decreasing group. The female tobacco decreasing group underscored not only the important role of parental knowledge and peer tobacco use but also the significance of victimization and peer alcohol use. The female alcohol decreasing group initially showed a similar level of alcohol use compared with the mid-stable group, but parental knowledge then significantly increased, compared with that of the mid-stable group, and the group then became less likely to associate with alcohol-using peers. In addition, victimization was initially linked to tobacco use, and then was sharply minimized, and peers’ alcohol use showed a consistent decreasing pattern for the female tobacco decreasing group. This is in line with explanations about the relationship between unstructured socialization and delinquency (Osgood, Wilson, O’Malley, Bachman, & Johnston, 1996). In congruence with previous research, shared beginnings but divergent outcomes are facilitated by structured socialization, which is closely related to parental knowledge and supervision (Deković, 1999; Dishion et al., 1995; Forehand et al., 1997; Li et al., 2000; Steinberg et al., 1994; Westling et al., 2008) and minimized peer pressure (Ary et al., 1993; Brook et al., 1990; Dupre et al., 1995; Elliott et al., 1985; Hawkins et al., 1992).

This study also found meaningful explanations for membership into different trajectory groups based on other diverse factors: First, low self-control was identified as a strong predictor for both alcohol and tobacco use patterns, especially for most prevalent groups. Moffitt’s (1993) typological approach provides specific explanations for those who are chronic offenders, versus those who are limited-time offenders during adolescence. Major differences between groups were found in propensity for antisocial behaviors, such as low self-control. All multinomial logistic regression models support such theoretical assumptions.

Second, gendered differences emerged from the role of peer attachment in substance use. Higher peer attachment serves as a protective factor for female tobacco use. Females who indicated higher levels of peer attachment were less likely to initiate tobacco use or to continue using it. However, higher levels of peer attachment appear to be a risk factor for alcohol use by males. Specifically, higher levels of peer attachment increased the risk of initiating alcohol usage. A possible explanation for the connection between attachment to peers and substance use comes from Akers (1998), who asserted that being attached to a deviant peer is a risk factor for delinquency, rather than a protective factor. In addition, Korean teenagers are exposed to their peers for a long time as they spend most of the day (12-15 hr) with them in school. Illegal behaviors, including substance use, may be helpful for boys who wish to be perceived as “cool kids” and avoid school bullying, a major social issue in South Korea. While girls may have the same needs, traditional gender roles are still strongly ingrained within Korean society. Although the country is becoming increasingly Westernized, Confucian values still persist in South Korean culture. Therefore, females are strongly encouraged to adhere to strict gender roles and tend to receive close parental supervision, whereas boys are allowed more freedom and tend to receive less attention from their parents (Jang & Jung, 2008; Y. Kim et al., 2001; Shin, 2004). Such cultural context may help explain gendered differences in the role of peer attachment.

Third, parental attachment emerged as a protective factor for both men and women but with different impacts on trajectory groups by gender. For women, parental attachment was associated with more habitual substance use groups. Alternatively, for men, parental attachment promoted membership into the lower prevalence groups. Prior research generally suggests that the impact of parental attachment on substance use appears to be stronger for women than for men (Bahr et al., 1995; Barfield-Cottledge, 2015; Farrell & White, 1998; Johnson & Marcos, 1988), but this study raised the possibility that the impact of parental attachment may vary based on the level of substance use by gender.

Differences in vulnerability to risk factors between males and females have often been discussed in criminology. For example, Moffitt and colleagues (2001) argued that

[m]ost of the risk factors for antisocial behaviour applied equally well to males and females in the Dunedin Study . . . Evaluated as a whole, then, the evidence suggests that sex differences in association between risk factors and antisocial behaviour are small and not very robust across different measurement strategies and statistical procedures. (p. 106)

Moffitt et al.’s (2001) comments were based on overall deviant behavior rather than substance use alone; however, they are still applicable to substance use and useful for understanding the link between gender and substance use.

Regarding gendered differences in antisocial behavior, Moffitt et al. (2001) hypothesized differential exposure to risk factors among males and females, which were partially supported. Consistent with Moffitt et al. (2001), Newcomb and colleagues (1986) and Nolen-Hoeksema (2004) found no gender gap in risk factors for substance use. Nolen-Hoeksema (2004) concluded, “[t]he bulk of the evidence suggests that relationships between risk factors and alcohol use or problems are more similar than different in women and men” (p. 993). This study highlights the necessity of research based on non-Western cultures for broadening our understanding of gender issues in substance use.

Although the current study found some support for similarities between genders in the link between certain predictors and substance use, we also found that different factors predicted membership in different alcohol and tobacco trajectory groups. This discrepancy may be due to the different methodological approach of the present study. Unlike Moffitt et al. (2001), Newcomb et al. (1986), and Nolen-Hoeksema (2004), this study employed a group-based trajectory analysis, which allowed for the exploration of subgroups based on the link between substance use and gender. Previous studies have focused on overall male and female groups (i.e., one male group and one female group), while the current study identified multiple latent groups for each gender (i.e., four male groups and four female groups) and compared them in detail.

Gender-role expectations between parents and children also need to be considered when examining gendered factors in substance use trajectory groups. The youth often attempt to forge independence from their parents during adolescence. Seeking autonomy may create conflict between youth and their parents. Such conflict may be more severe for males than for females because of males’ attitudes toward parents, and the fact that males’ family norms tend to be more confrontational compared with that of females (Kort-Butler, 2009; Whitesell, Bachand, Peel, & Brown, 2013). This conflict may weaken parental knowledge and supervision of boys as parents and children influence one another (Bronfenbrenner, 2001; Cho & Yoon, 2010).

In addition to the theoretical contributions of the current study, our findings have implications for policy. Enhancing protective factors and reducing risk factors for substance use are major principles of many prevention efforts (Hawkins, Catalano, & Arthur, 2002; Robertson, David, & Rao, 2003). Therefore, by investigating various factors in different domains (individual, parental, and peer domains) for various substance use groups, this study has particular relevance for prevention policy. Results of the present study show the salient effects of parental and peer factors. To be specific, parental knowledge of youth whereabouts and activities appears to function as a protective factor, whereas peer substance use is a risk factor for substance use. Therefore, prevention strategies should include parent-targeted programs which provide parents with knowledge and training on rule-setting, setting clear expectations, and skills for supervising activities as well as disciplining children. In addition, peer counseling, peer coping/resistance strategies, promoting positive peer engagement, and enhancing responsible choices could modify and reduce the influence of substance-using peers. In light of better performance of gender-responsive substance use treatment (Messina, Grella, Cartier, & Torres, 2010), gender-specific strategies are also an important pursuit. Enhancing parental attachment should be targeted at men in the nonuse categories to prevent initiation and women who are already using substances to minimize their continued consumption.

Our study is not without limitations. First, this study only analyzed five waves in late adolescence due to data limitations. Although the current analysis relied on data which only included youth aged 14 to 18, this age is a very critical period in the development of young people. This age span covers early (in part), middle, and late adolescence, a time when youth are experiencing many physical, psychological, and social development changes (American Psychological Association, 2002; Christie & Viner, 2005; Petersen, 1988). Criminological theories also assert propositions regarding the importance of changes during this time period in influencing behavior over the life course. For instance, Moffitt (1993) theorized a typological approach which consists of adolescence-limited and life-course-persistent offenders, and suggested that the development and progression of deviant behaviors of these two groups are closely related to middle and late adolescence. Analyzing data from more waves, including early childhood and adulthood, may allow researchers to investigate extended trajectories for males and females. Such an examination would also allow for exploration of long-term effects of risk and protective factors on substance use.

Second, more research should explore the female decreasing trajectory group. Although this study identified various risk and protective factors of the female decreasing group, future research is needed which draws on different data sources to extend the current findings to more general notions. In particular, follow-up research is necessary to clarify more details on the background of gendered peer pressure dynamics. This will further illuminate our understanding of the connections between peer pressure and substance use.

Footnotes

Appendix

Correlation Matrix

Var.	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21
1	—
2	.40*	—
3	.31*	.42*	—
4	.24*	.37*	.52*	—
5	.23*	.34*	.40*	.45*	—
6	.32*	.27*	.27*	.26*	.20*	—
7	.27*	.39*	.35*	.34*	.27*	.49*	—
8	.19*	.33*	.42*	.43*	.32*	.47*	.56*	—
9	.16*	.31*	.40*	.47*	.36*	.40*	.51*	.67*	—
10	.18*	.30*	.35*	.36*	.41*	.35*	.45*	.54*	.66*	—
11	.28*	.23*	.25*	.23*	.21*	.24*	.23*	.23*	.22*	.25*	—
12	.18*	.13*	.12*	.11*	.11*	.23*	.12*	.18*	.10*	.10*	.14*	—
13	−.14*	−.07*	−.10*	−.08*	−.09*	−.12*	−.11*	−.09*	−.09*	−.09*	−.17*	−.05	—
14	−.20*	−.12*	−.14*	−.10*	−.14*	−.14*	−.13*	−.13*	−.11*	−.12*	−.28*	−.12*	.44*	—
15	.15*	.07*	.12*	.13*	.10*	.17*	.12*	.12*	.10*	.12*	.26*	.20*	−.24*	−.22*	—
16	.05	.08*	.05	.07*	.05	.08*	.07*	.05	.04	.04	−.07*	.00	.09*	.09*	−.05	—
17	.47*	.37*	.29*	.28*	.27*	.38*	.36*	.30*	.25*	.26*	.25*	.10*	−.09*	−.16*	.04	.06*	—
18	.31*	.26*	.23*	.21*	.23*	.53*	.41*	.34*	.26*	.27*	.19*	.13*	−.08*	−.14*	.08*	.04	.70*	—
19	−.03	−.01	−.03	−.07*	−.06*	−.06*	−.01	−.04	−.05	−.05	−.10*	−.06*	.09*	.14*	−.06*	.09*	−.03	−.03	—
20	−.01	.02	−.04	−.04	−.04	−.04	.00	−.01	−.03	−.01	−.05	−.04	.11*	.11*	−.03	.10*	−.03	−.03	.68*	—
21	−.01	.00	−.05	−.07*	−.03	−.05	−.04	−.08*	−.08*	−.08*	−.04	−.04	.04	.06*	−.01	.08*	−.03	.01	.37*	.38*	—

Note. Var. = Variable. 1 = alcohol use (Wave 1), 2 = alcohol use (Wave 2), 3 = alcohol use (Wave 3), 4 = alcohol use (Wave 4), 5 = alcohol use (Wave 5), 6 = tobacco use (Wave 1), 7 = tobacco use (Wave 2), 8 = tobacco use (Wave 3), 9 = tobacco use (Wave 4), 10 = tobacco use (Wave 5), 11 = low self-control, 12 = victimization, 13 = parental attachment, 14 = parental knowledge, 15 = parental abuse, 16 = peer attachment, 17 = peer alcohol use, 18 = peer tobacco use, 19 = paternal education level, 20 = maternal education level, 21 = family income.

p < .05.

Acknowledgements

We would like to thank Dr. Abigail A. Fagan and anonymous reviewers for their insightful comments on our draft.

Yeungjeom Lee is a criminology doctoral student in the Department of Sociology and Criminology & Law at the University of Florida. Her primary interests include life-course/developmental criminology and juvenile delinquency. Her most recent publications appear in Crime & Delinquency and Deviant Behavior.

Jihoon Kim is a doctoral candidate in the Department of Sociology and Criminology & Law at the University of Florida. His research interests lie in the area of life-course/developmental theory and quantitative analysis. His recent or forthcoming publications appear in Crime & Delinquency and Deviant Behavior.

References

Akers

R. L.

(1998). Social learning and social structure: A general theory of crime and deviance. Boston, MA: Northeastern University Press.

Akers

R. L.

Cochran

J. K.

(1985). Adolescent marijuana use: A test of three theories of deviant behavior. Deviant Behavior, 6, 323-346.

Akers

R. L.

Krohn

M. D.

Lanza-Kaduce

Radosevich

(1979). Social learning and deviant behavior: A specific test of a general theory. American Sociological Review, 44, 636-655.

American Psychological Association. (2002). Developing adolescents: A reference for professionals. Washington, DC: American Psychological Society.

Anthony

J. C.

Warner

L. A.

Kessler

R. C.

(1994). Comparative epidemiology of dependence on tobacco, alcohol, controlled substances, and inhalants: Basic findings from the national comorbidity survey. Experimental and Clinical Psychopharmacology, 2, 244-268.

Arseneault

Cannon

Poulton

Murray

Caspi

Moffitt

T. E.

(2002). Cannabis use in adolescence and risk for adult psychosis: Longitudinal prospective study. British Medical Journal, 325, 1212-1213.

Ary

D. V.

Tildesley

Hops

Andrews

(1993). The influence of parent, sibling, and peer modeling and attitudes on adolescent use of alcohol. International Journal of the Addictions, 28, 853-880.

Bahr

S. J.

Marcos

A. C.

Maughan

S. L.

(1995). Family, educational and peer influences on the alcohol use of female and male adolescents. Journal of Studies on Alcohol, 56, 457-469.

Barfield-Cottledge

(2015). The triangulation effects of family structure and attachment on adolescent substance use. Crime & Delinquency, 61, 297-320.

10.

Baumrind

(1985). Familial antecedents of adolescent drug use: A developmental perspective. In Jones

C. L.

Battjes

R. J.

(Eds.), Etiology of drug abuse: Implications for prevention (NIDA Research Monograph No. 56, DHHS Publication No ADM. 85-1355, pp. 13-44). Rockville, MD: National Institute on Drug Abuse.

11.

Bennett

Holloway

Farrington

(2008). The statistical association between drug misuse and crime: A meta-analysis. Aggression and Violent Behavior, 13, 107-118.

12.

Berg

Kiviruusu

Karvonen

Kestilä

Lintonen

Rahkonen

Huurre

(2013). A 26-year follow-up study of heavy drinking trajectories from adolescence to mid-adulthood and adult disadvantage. Alcohol and Alcoholism, 48, 452-457.

13.

Biehl

M. C.

Natsuaki

M. N.

(2007). The influence of pubertal timing on alcohol use and heavy drinking trajectories. Journal of Youth and Adolescence, 36, 153-167.

14.

Brady

K. T.

Randall

C. L.

(1999). Gender differences in substance use disorders. Psychiatric Clinics of North America, 22, 241-252.

15.

Bronfenbrenner

(2001). Human development, bioecological theory of. In Smelser

N. J.

Baltes

P. B.

(Eds.), International encyclopedia of the social and behavioral sciences (pp. 6963-6970). Oxford, UK: Elsevier.

16.

Brook

J. S.

Brook

D. W.

Gordon

A. S.

Whiteman

Cohen

(1990). The psychosocial etiology of adolescent drug use: A family interactional approach. Genetic, Social, and General Psychology Monographs, 116, 111-267.

17.

Brook

J. S.

Gordon

A. S.

Brook

D. W.

(1989). The consequences of marijuana use on intrapersonal and interpersonal functioning in black and white adolescents. Genetic, Social, and General Psychology Monographs, 115, 349-369.

18.

Brook

J. S.

Whiteman

Gordon

A. S.

(1982). Qualitative and quantitative aspects of adolescent drug use: Interplay of personality, family, and peer correlates. Psychological Reports, 51, 1151-1163.

19.

Brook

J. S.

Whiteman

Gordon

A. S.

Brook

D. W.

(1984). Paternal determinants of female adolescent’s marijuana use. Developmental Psychology, 20, 1032-1043.

20.

Buu

Dabrowska

Heinze

J. E.

Hsieh

H. F.

Zimmerman

M. A.

(2015). Gender differences in the developmental trajectories of multiple substance use and the effect of nicotine and marijuana use on heavy drinking in a high-risk sample. Addictive Behaviors, 50, 6-12.

21.

Caspi

Begg

Dickson

Harrington

Langley

Moffitt

T. E.

Silva

P. A.

(1997). Personality differences predict health-risk behaviors in young adulthood: Evidence from a longitudinal study. Journal of Personality and Social Psychology, 73, 1052-1063.

22.

Caspi

Moffitt

T. E.

Newman

D. L.

Silva

P. A.

(1996). Behavioral observations at age 3 years predict adult psychiatric disorders: Longitudinal evidence from a birth cohort. Archives of General Psychiatry, 53, 1033-1039.

23.

Cernkovich

S. A.

Giordano

P. C.

(1987). Family relationships and delinquency. Criminology, 25, 295-319.

24.

Chen

Jacobson

K. C.

(2012). Developmental trajectories of substance use from early adolescence to young adulthood: Gender and racial/ethnic differences. Journal of Adolescent Health, 50, 154-163.

25.

Chi

Lubben

J. E.

Kitano

H. H.

(1989). Differences in drinking behavior among three Asian-American groups. Journal of Studies on Alcohol, 50, 15-23.

26.

Cho

Yoon

(2010). A longitudinal study of reciprocal relation between parental monitoring and adolescent drinking. Korean Journal of Youth Studies, 17, 259-285.

27.

Christie

Viner

(2005). Adolescent development. British Medical Journal, 330, 301-304.

28.

Chun

Chung

I. J.

(2013). Gender differences in factors influencing smoking, drinking, and their co-occurrence among adolescents in South Korea. Nicotine & Tobacco Research, 15, 542-551.

29.

Cloninger

C. R.

Sigvardsson

Bohman

(1988). Childhood personality predicts alcohol abuse in young adults. Alcoholism: Clinical & Experimental Research, 12, 494-505.

30.

Coombs

R. H.

Paulson

M. J.

Richardson

M. A.

(1991). Peer vs. parental influence in substance use among Hispanic and Anglo children and adolescents. Journal of Youth and Adolescence, 20, 73-88.

31.

Dawson

D. A.

T. K.

Grant

B. F.

(2008). A prospective study of risk drinking: At risk for what? Drug and Alcohol Dependence, 95, 62-72.

32.

Deković

(1999). Risk and protective factors in the development of problem behavior during adolescence. Journal of Youth and Adolescence, 28, 667-685.

33.

DeWit

D. J.

Adlaf

E. M.

Offord

D. R.

Ogborne

A. C.

(2000). Age at first alcohol use: A risk factor for the development of alcohol disorders. American Journal of Psychiatry, 157, 745-750.

34.

Dishion

T. J.

Capaldi

Spracklen

K. M.

(1995). Peer ecology of male adolescent drug use. Development and Psychopathology, 7, 803-824.

35.

Dishion

T. J.

Loeber

(1985). Adolescent marijuana and alcohol use: The role of parents and peers revisited. The American Journal of Drug and Alcohol Abuse, 11, 11-25.

36.

Dishion

T. J.

Nelson

S. E.

Bullock

B. M.

(2004). Premature adolescent autonomy: Parent disengagement and deviant peer process in the amplification of problem behaviour. Journal of Adolescence, 27, 515-530.

37.

Dishion

T. J.

Patterson

G. R.

Reid

J. R.

(1988). Parent and peer factors associated with drug sampling in early adolescence: Implications for treatment. NIDA Research Monograph, 77, 69-93.

38.

Duncan

S. C.

Alpert

Duncan

T. E.

Hops

(1997). Adolescent alcohol use development and young adult outcomes. Drug and Alcohol Dependence, 49, 39-48.

39.

Duncan

S. C.

Duncan

T. E.

Strycker

L. A.

(2006). Alcohol use from ages 9 to 16: A cohort-sequential latent growth model. Drug and Alcohol Dependence, 81, 71-81.

40.

Duncan

T. E.

Duncan

S. C.

Hops

(1994). The effects of family cohesiveness and peer encouragement on the development of adolescent alcohol use: A cohort-sequential approach to the analysis of longitudinal data. Journal of Studies on Alcohol, 55, 588-599.

41.

Dupre

Miller

Gold

Rospenda

(1995). Initiation and progression of alcohol, marijuana, and cocaine use among adolescent abusers. The American Journal on Addictions, 4, 43-48.

42.

Ellickson

P. L.

Hays

R. D.

Bell

R. M.

(1992). Stepping through the drug use sequence: Longitudinal scalogram analysis of initiation and regular use. Journal of Abnormal Psychology, 101, 441-451.

43.

Elliott

D. S.

Huizinga

Ageton

S. S.

(1985). Explaining delinquency and drug use. Beverly Hills, CA: SAGE.

44.

Evans-Polce

R. J.

Vasilenko

S. A.

Lanza

S. T.

(2015). Changes in gender and racial/ethnic disparities in rates of cigarette use, regular heavy episodic drinking, and marijuana use: Ages 14 to 32. Addictive Behaviors, 41, 218-222.

45.

Exum

M. L.

(2006). Alcohol and aggression: An integration of findings from experimental studies. Journal of Criminal Justice, 34, 131-145.

46.

Fagan

A. A.

Van Horn

M. L.

Hawkins

J. D.

Jaki

(2013). Differential effects of parental controls on adolescent substance use: For whom is the family most important? Journal of Quantitative Criminology, 29, 347-368.

47.

Farrell

A. D.

White

K. S.

(1998). Peer influences and drug use among urban adolescents: Family structure and parent–adolescent relationship as protective factors. Journal of Consulting and Clinical Psychology, 66, 248-258.

48.

Flannery

D. J.

Vazsonyi

A. T.

Torquati

Fridrich

(1994). Ethnic and gender differences in risk for early adolescent substance use. Journal of Youth and Adolescence, 23, 195-213.

49.

Forehand

Miller

K. S.

Dutra

Chance

M. W.

(1997). Role of parenting in adolescent deviant behavior: Replication across and within two ethnic groups. Journal of Consulting and Clinical Psychology, 65, 1036-1041.

50.

Getting

E. R.

Beauvais

(1986). Peer cluster theory: Drugs and the adolescent. Journal of Counseling and Development, 65, 17-22.

51.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Stanford, CA: Stanford University Press.

52.

Grant

B. F.

Dawson

D. A.

(1998). Age of onset of drug use and its association with DSM-IV drug abuse and dependence: Results from the national longitudinal alcohol epidemiologic survey. Journal of Substance Abuse, 10, 163-173.

53.

Grasmick

H. G.

Tittle

C. R.

Bursik

R. J.

Arneklev

B. J.

(1993). Testing the core empirical implications of Gottfredson and Hirschi’s general theory of crime. Journal of Research in Crime & Delinquency, 30, 5-29.

54.

Griffin

K. W.

Botvin

G. J.

Scheier

L. M.

Diaz

Miller

N. L.

(2000). Parenting practices as predictors of substance use, delinquency, and aggression among urban minority youth: Moderating effects of family structure and gender. Psychology of Addictive Behaviors, 14, 174-184.

55.

Han

Grogan-Kaylor

(2013). Parenting and youth psychosocial well-being in South Korea using fixed-effects models. Journal of Family Issues, 34, 689-715.

56.

Hawkins

J. D.

Catalano

R. F.

Arthur

M. W.

(2002). Promoting science-based prevention in communities. Addictive Behaviors, 27, 951-976.

57.

Hawkins

J. D.

Catalano

R. F.

Miller

J. Y.

(1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin, 112, 64-105.

58.

Hawkins

J. D.

Graham

J. W.

Maguin

Abbott

Hill

K. G.

Catalano

R. F.

(1997). Exploring the effects of age of alcohol use initiation and psychosocial risk factors on subsequent alcohol misuse. Journal of Studies on Alcohol, 58, 280-290.

59.

Hesselbrock

M. N.

Meyer

R. E.

Keener

J. J.

(1985). Psychopathology in hospitalized alcoholics. Archives of General Psychiatry, 42, 1050-1055.

60.

Hill

K. G.

White

H. R.

Chung

I. J.

Hawkins

J. D.

Catalano

R. F.

(2000). Early adult outcomes of adolescent binge drinking: Person-and variable-centered analyses of binge drinking trajectories. Alcoholism, Clinical & Experimental Research, 24, 892-901.

61.

Hops

Davis

B. L.

Lewin

L. M.

(1999). The development of alcohol and other substance use: A gender study of family and peer context. Journal of Studies on Alcohol, Supplement, 13, 22-31.

62.

Hops

Tildesley

Lichtenstein

Ary

Sherman

(1990). Parent-adolescent problem-solving interactions and drug use. The American Journal of Drug and Alcohol Abuse, 16, 239-258.

63.

Hwang

(2010). Parental and peer factors influence on adolescents’ drinking and smoking behaviors. Health and Social Science, 27, 81-103.

64.

Hwang

(2012). The approach of theories of deviant behavior on adolescents’ smoking and drinking behaviors based on general strain theory and social learning theory. Health and Social Science, 32, 19-38.

65.

Hwang

Akers

R. L.

(2003). Substance use by Korean adolescents: A cross-cultural test of social learning, social bonding, and self-control theories. In Akers

R. L.

Jensen

G. F.

(Eds.), Social learning and the explanation of crime: A guide for the new century, advances in criminological theory (Vol. 11, pp. 39-63). New Brunswick, NJ: Transaction.

66.

Hwang

Akers

R. L.

(2006). Parental and peer influences on adolescent drug use in Korea. Asian Journal of Criminology, 1, 51-69.

67.

Jacob

Johnson

S. L.

(1997). Parent–child interaction among depressed fathers and mothers: Impact on child functioning. Journal of Family Psychology, 11, 391-409.

68.

Jang

S. O.

Jung

(2008). Sex differences in drinking behavior and sex role expectations. Korean Journal of Social Sciences, 27, 89-106.

69.

Jennison

K. M.

(2004). The short-term effects and unintended long-term consequences of binge drinking in college: A 10-year follow-up study. The American Journal of Drug and Alcohol Abuse, 30, 659-684.

70.

Jessor

S. L.

(1977). Problem behavior and psychosocial development: A longitudinal study of youth. New York, NY: Academic Press.

71.

Johnson

R. E.

Marcos

A. C.

(1988). Correlates of adolescent drug use by gender and geographic location. The American Journal of Drug and Alcohol Abuse, 14, 51-63.

72.

Johnston

L. D.

O’Malley

P. M.

Bachman

J. G.

Schulenberg

J. E.

(2011). Monitoring the future national results on adolescent drug use: Overview of key findings, 2010. Ann Arbor: Institute for Social Research, University of Michigan.

73.

Johnston

L. D.

O’Malley

P. M.

Bachman

J. G.

Schulenberg

J. E.

(2013). Monitoring the future national survey results on drug use, 1975-2012: Volume I, Secondary school students. Ann Arbor: Institute for Social Research, University of Michigan.

74.

Jones

B. L.

Nagin

D. S.

(2012). A Stata plugin for estimating group-based trajectory models. Heinz College Research, Carnegie Mellon University. Retrieved from https://www.andrew.cmu.edu/~bjones/index.htm

75.

Kandel

D. B.

Yamaguchi

Chen

(1992). Stages of progression in drug involvement from adolescence to adulthood: Further evidence for the gateway theory. Journal of Studies on Alcohol, 53, 447-457.

76.

Kaufman

J. M.

(2009). Gendered responses to serious strain: The argument for a general strain theory of deviance. Justice Quarterly, 26, 410-444.

77.

Kerr

Stattin

(2000). What parents know, how they know it, and several forms of adolescent adjustment: Further support for a reinterpretation of monitoring. Developmental Psychology, 36, 366-380.

78.

Kessler

R. C.

Crum

R. M.

Warner

L. A.

Nelson

C. B.

Schulenberg

Anthony

J. C.

(1997). Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the national comorbidity survey. Archives of General Psychiatry, 54, 313-321.

79.

Kilpatrick

D. G.

Acierno

Saunders

Resnick

H. S.

Best

C. L.

Schnurr

P. P.

(2000). Risk factors for adolescent substance abuse and dependence: Data from a national sample. Journal of Consulting and Clinical Psychology, 68, 19-30.

80.

Kim

Kwak

D. H.

Yun

(2010). Investigating the effects of peer association and parental influence on adolescent substance use: A study of adolescents in South Korea. Journal of Criminal Justice, 38, 17-24.

81.

Kim

(2014). Predictors of drinking behavior in Korean male and female youth. Korean Journal of Woman Psychology, 19, 191-210.

82.

Kim

(2001). A national study on adolescent alcohol use: Prevalence and related factors. Korean Journal of Social Welfare, 47, 71-106.

83.

Kippin

T. E.

Fuchs

R. A.

Mehta

R. H.

Case

J. M.

Parker

M. P.

Bimonte-Nelson

H. A.

See

R. E.

(2005). Potentiation of cocaine-primed reinstatement of drug seeking in female rats during estrus. Psychopharmacology, 182, 245-252.

84.

Korean Association Against Drug Abuse. (2008). Survey on youth awareness of the severity of drugs in 2007. Seoul: Author.

85.

Korean Ministry for Health Welfare and Family Affairs & Korean Association of Smoking and Health. (2008). The reality of smoking. Seoul: Korean Ministry for Health Welfare and Family Affairs.

86.

Korean Ministry of Gender Equality and Family. (2012). Study on adolescent exposure to harmful environment (Publication No. 11-1383000-000015-10). Seoul: Author.

87.

Korean National Youth Policy Institute. (2010). Survey on Korean adolescent health. Seoul: Author.

88.

Kort-Butler

L. A.

(2009). Coping styles and sex differences in depressive symptoms and delinquent behavior. Journal of Youth and Adolescence, 38, 122-136.

89.

Lee

(2012). The influence of parent–adolescent relationship quality, parental monitoring, and peer substance use on substance use in South Korean high school students: Using latent growth curve modelling. Health Education Journal, 71, 505-515.

90.

Lee

Kim

(2016). An examination of victimization trajectories among a sample of South Korean adolescents: Risk and protective factors. Crime & Delinquency. Advance online publication. doi:10.1177/0011128716634103

91.

Feigelman

Stanton

(2000). Perceived parental monitoring and health risk behaviors among urban low-income African-American children and adolescents. Journal of Adolescent Health, 27, 43-48.

92.

Loeber

Dishion

(1983). Early predictors of male delinquency: A review. Psychological Bulletin, 94, 68-99.

93.

Marshal

M. P.

(2003). For better or for worse? The effects of alcohol use on marital functioning. Clinical Psychology Review, 23, 959-997.

94.

Martens

P. L.

(1997). Parental monitoring and deviant behaviour among juveniles. Studies on Crime & Crime Prevention, 6, 224-244.

95.

Messina

Grella

C. E.

Burdon

Prendergast

(2007). Childhood adverse events and current traumatic distress a comparison of men and women drug-dependent prisoners. Criminal Justice and Behavior, 34, 1385-1401.

96.

Messina

Grella

C. E.

Cartier

Torres

(2010). A randomized experimental study of gender-responsive substance abuse treatment for women in prison. Journal of Substance Abuse Treatment, 38, 97-107.

97.

Moffitt

T. E.

(1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674-701.

98.

Moffitt

T. E.

Caspi

Rutter

Silva

P. A.

(2001). Sex differences in antisocial behaviour: Conduct disorder, delinquency, and violence in the Dunedin Longitudinal Study. Cambridge, UK: Cambridge University Press.

99.

Moran

P. B.

Vuchinich

Hall

N. K.

(2004). Associations between types of maltreatment and substance use during adolescence. Child Abuse & Neglect, 28, 565-574.

100.

Mullahy

Sindelar

(1996). Employment, unemployment, and problem drinking. Journal of Health Economics, 15, 409-434.

101.

Muthén

B. O.

Muthén

L. K.

(2000). The development of heavy drinking and alcohol-related problems from ages 18 to 37 in a U.S. national sample. Journal of Studies on Alcohol, 61, 290-300.

102.

Nagin

D. S.

(2005). Group-based modeling of development. Cambridge, MA: Harvard University Press.

103.

Nelson

S. E.

Van Ryzin

M. J.

Dishion

T. J.

(2015). Alcohol, marijuana, and tobacco use trajectories from age 12 to 24 years: Demographic correlates and young adult substance use problems. Development and Psychopathology, 27, 253-277.

104.

Newcomb

M. D.

Bentler

P. M.

(1987). The impact of late adolescent substance use on young adult health status and utilization of health services: A structural-equation model over four years. Social Science & Medicine, 24, 71-82.

105.

Newcomb

M. D.

Bentler

P. M.

(1988). Consequences of adolescent drug use: Impact on the lives of young adults. Beverly Hills, CA; SAGE.

106.

Newcomb

M. D.

Maddahian

Bentler

P. M.

(1986). Risk factors for drug use among adolescents: Concurrent and longitudinal analyses. American Journal of Public Health, 76, 525-531.

107.

Nolen-Hoeksema

(2004). Gender differences in risk factors and consequences for alcohol use and problems. Clinical Psychology Review, 24, 981-1010.

108.

Osgood

D. W.

Wilson

J. K.

O’Malley

P. M.

Bachman

J. G.

Johnston

L. D.

(1996). Routine activities and individual deviant behavior. American Sociological Review, 61, 635-655.

109.

Özbay

. (2008). Does social capital deter youth from cheating, alcohol use, and violence in Turkey? Bringing torpil in. Journal of Criminal Justice, 36, 403-415.

110.

Park

(2007). Factors affecting cigarette use and an increase in smoking frequency among adolescents in South Korea. Child Health Nursing Research, 13, 318-328.

111.

Paternoster

Triplett

(1988). Disaggregating self-reported delinquency and its implications for theory. Criminology, 26, 591-626.

112.

Petersen

A. C.

(1988). Adolescent development. Annual Review of Psychology, 39, 583-607.

113.

Pinchevsky

G. M.

Wright

E. M.

Fagan

A. A.

(2013). Gender differences in the effects of exposure to violence on adolescent substance use. Violence and Victims, 28, 122-144.

114.

Plant

Miller

Thornton

Plant

Bloomfield

(2000). Life stage, alcohol consumption patterns, alcohol-related consequences, and gender. Substance Abuse, 21, 265-281.

115.

Pollard

J. A.

Hawkins

J. D.

Arthur

M. W.

(1999). Risk and protection: Are both necessary to understand diverse behavioral outcomes in adolescence? Social Work Research, 23, 145-158.

116.

Pratt

T. C.

Cullen

F. T.

(2000). The empirical status of Gottfredson and Hirschi’s general theory of crime: A meta-analysis. Criminology, 38, 931-964.

117.

Robertson

E. B.

David

S. L.

Rao

S. A.

(2003). Preventing drug abuse among children and adolescents: A research-based guide for parents, educators, and community leaders (NIH Publication No. 04-4212[A]). Bethesda, MD: U.S. National Institutes of Health, National Institute on Drug Abuse.

118.

Rodgers

C. S.

Lang

A. J.

Laffaye

Satz

L. E.

Dresselhaus

T. R.

Stein

M. B.

(2004). The impact of individual forms of childhood maltreatment on health behavior. Child Abuse & Neglect, 28, 575-586.

119.

Roerecke

Rehm

(2013). Alcohol use disorders and mortality: A systematic review and meta-analysis. Addiction, 108, 1562-1578.

120.

Rutledge

P. C.

Sher

K. J.

(2001). Heavy drinking from the freshman year into early young adulthood: The roles of stress, tension-reduction drinking motives, gender and personality. Journal of Studies on Alcohol, 62, 457-466.

121.

Schramm-Sapyta

N. L.

Walker

Q. D.

Caster

J. M.

Levin

E. D.

Kuhn

C. M.

(2009). Are adolescents more vulnerable to drug addiction than adults? Evidence from animal models. Psychopharmacology, 206, 1-21.

122.

Sher

K. J.

Trull

T. J.

Bartholow

B. D.

Vieth

(1999). Personality and alcoholism: Issues, methods, and etiological processes. In Leonard

Blane

H. T.

(Eds.), Psychological theories on drinking and alcoholism (pp. 54-105). New York, NY: Guildford Press.

123.

Shin

H. W.

(2004). A study on the relationship between hyperfemininity and alcohol use in undergraduate female students. The Korean Journal of Health Psychology, 9, 693-711.

124.

Slutske

W. S.

Heath

A. C.

Madden

P. A.

Bucholz

K. K.

Statham

D. J.

Martin

N. G.

(2002). Personality and the genetic risk for alcohol dependence. Journal of Abnormal Psychology, 111, 124-133.

125.

Smith

D. A.

Paternoster

(1987). The gender gap in theories of deviance: Issues and evidence. Journal of Research in Crime & Delinquency, 24, 140-172.

126.

Stattin

Kerr

(2000). Parental monitoring: A reinterpretation. Child Development, 71, 1072-1085.

127.

Steinberg

Fletcher

Darling

(1994). Parental monitoring and peer influences on adolescent substance use. Pediatrics, 93, 1060-1064.

128.

Steinberg

Silverberg

S. B.

(1986). The vicissitudes of autonomy in early adolescence. Child Development, 57, 841-851.

129.

The Substance Abuse and Mental Health Services Administration. (2012). Results from the 2010 national survey on drug use and health: Mental health findings (NSDUH series H-42, HHS publication no. (SMA) 11-4667). Rockville, MD: Author.

130.

Svensson

(2003). Gender differences in adolescent drug use the impact of parental monitoring and peer deviance. Youth & Society, 34, 300-329.

131.

Thornberry

T. P.

(1987). Toward an interactional theory of delinquency. Criminology, 25, 863-892.

132.

Thornberry

T. P.

Krohn

M. D.

(1997). Peers, drug use, and delinquency. In Stoff

D. M.

Breiling

Maser

J. D.

(Eds.), Handbook of antisocial behavior (pp. 218-233). New York, NY: John Wiley.

133.

Tucker

J. S.

Ellickson

P. L.

Orlando

Martino

S. C.

Klein

D. J.

(2005). Substance use trajectories from early adolescence to emerging adulthood: A comparison of smoking, binge drinking, and marijuana use. Journal of Drug Issues, 35, 307-332.

134.

Waldeck

T. L.

Miller

L. S.

(1997). Gender and impulsivity differences in licit substance use. Journal of Substance Abuse, 9, 269-275.

135.

Warr

(2002). Companions in crime: The social aspects of criminal conduct. New York, NY: Cambridge University Press.

136.

Wechsler

Davenport

Dowdall

Moeykens

Castillo

(1994). Health and behavioral consequences of binge drinking in college: A national survey of students at 140 campuses. Journal of American Medical Association, 272, 1672-1677.

137.

Weintraub

K. J.

Gold

(1991). Monitoring and delinquency. Criminal Behaviour and Mental Health, 1, 268-281.

138.

Westling

Andrews

J. A.

Hampson

S. E.

Peterson

(2008). Pubertal timing and substance use: The effects of gender, parental monitoring and deviant peers. Journal of Adolescent Health, 42, 555-563.

139.

Whaley

R. B.

Hayes-Smith

(2013). Gendered pathways? Gender, mediating factors, and the gap in boys’ and girls’ substance use. Crime & Delinquency, 59, 651-669.

140.

White

H. R.

Johnson

Buyske

(2000). Parental modeling and parenting behavior effects on offspring alcohol and cigarette use: A growth curve analysis. Journal of Substance Abuse, 12, 287-310.

141.

Whitesell

Bachand

Peel

Brown

(2013). Familial, social, and individual factors contributing to risk for adolescent substance use. Journal of Addiction, 2013, Article 579310. doi:10.1155/2013/579310

142.

Wills

T. A.

Windle

Cleary

S. D.

(1998). Temperament and novelty seeking in adolescent substance use: Convergence of dimensions of temperament with constructs from Cloninger’s theory. Journal of Personality and Social Psychology, 74, 387-406.

143.

Windle

Mun

E. Y.

Windle

R. C.

(2005). Adolescent-to-young adulthood heavy drinking trajectories and their prospective predictors. Journal of Studies on Alcohol, 66, 313-322.

144.

Winfree

L. T.

Jr. Griffiths

C. T.

(1983). Social learning and adolescent marijuana use: A trend study of deviant behavior in a rural middle school. Rural Sociology, 48, 219-239.

145.

(1998). Alcohol consumption and breast cancer risk. Journal of American Medical Association, 280, 1138-1139.

146.

Zhang

Wieczorek

W. F.

Welte

J. W.

Colder

Nochajski

T. H.

(2010). Delinquency and alcohol-impaired driving among young men: A longitudinal study. Journal of Criminal Justice, 38, 439-445.

147.

Zuckerman

Kuhlman

D. M.

(2000). Personality and risk-taking: Common biosocial factors. Journal of Personality, 68, 999-1029.