Parenting Style and Behavior as Longitudinal Predictors of Adolescent ...

Download PDF
2 downloads 0 Views 111KB Size Report
Comment
The present study tested Baumrind's proposal that parenting styles are direct predictors of adolescent alcohol use. Method: Latent class modeling was used to.
MINAIE ET AL.

671

Parenting Style and Behavior as Longitudinal Predictors of Adolescent Alcohol Use MATIN GHAYOUR MINAIE, PH.D.,a KA KIT HUI, B.A.,a RACHEL K. LEUNG, PH.D.,a,b,c,* JOHN W. TOUMBOUROU, PH.D.,a,b,c & ROSS M. KING, PH.D.a aSchool

of Psychology, Deakin University, Geelong, Victoria, Australia for Social and Early Emotional Development, Deakin University, Geelong, Victoria, Australia cCentre for Adolescent Health, Murdoch Children’s Research Institute, Parkville, Victoria, Australia bCentre

ABSTRACT. Objective: Adolescent alcohol use is a serious problem in Australia and other nations. Longitudinal data on family predictors are valuable to guide parental education efforts. The present study tested Baumrind’s proposal that parenting styles are direct predictors of adolescent alcohol use. Method: Latent class modeling was used to investigate adolescent perceptions of parenting styles and multivariate regression to examine their predictive effect on the development of adolescent alcohol use. The data set comprised 2,081 secondary school students (55.9% female) from metropolitan Melbourne, Australia, who

completed three waves of annual longitudinal data starting in 2004. Results: Baumrind’s parenting styles were significant predictors in unadjusted analyses, but these effects were not maintained in multivariate models that also included parenting behavior dimensions. Conclusions: Family influences on the development of adolescent alcohol use appear to operate more directly through specific family management behaviors rather than through more global parenting styles. (J. Stud. Alcohol Drugs, 76, 671–679, 2015)

A

Several studies have specifically investigated the influence of parents and families on adolescent alcohol use (Bahr & Hoffmann, 2010; Clausen, 1996; Koning et al., 2013; Mares et al., 2012). A number of parental behaviors influence adolescent substance use; for instance, the risk of adolescent alcohol and other drug use increases with parental provision of alcohol (Shortt et al., 2007), inconsistent parental feedback, and unreasonably severe punishment (Brook et al., 1990). The risk also rises when adolescents perceive a lack of closeness, warmth, and involvement with parents (Ryan et al., 2010). Large longitudinal studies examining the impact of parental demanding (e.g., control efforts, setting and monitoring rules) and nurturing (e.g., affection, support) behaviors report that both contribute independently and directly to the prediction of adolescent alcohol use (e.g., Huh et al., 2006; Latendresse et al., 2008; Pires & Jenkins, 2007). Four studies suggest that parental nurturance and/or parental demandingness affect adolescent alcohol use indirectly by mediating factors such as parental monitoring, self-regulation, genetic vulnerability, and peers (Barnes et al., 2000; Brody & Ge, 2001; Brody et al., 2009; Coker & Borders, 2001). A number of discrepancies can be found between the findings of the reviewed longitudinal studies examining the effect of parenting behaviors on adolescent alcohol use. Some studies have found a significant effect of parental monitoring but a nonsignificant effect of parental warmth in predicting adolescent alcohol use (Aquilino & Supple, 2001; Huh et al., 2006; Koning et al., 2012; Mares et al., 2012). A smaller number of studies found parental warmth but not parental monitoring to be predictive (Coker & Borders, 2001; Stice et al., 1998) or that neither of these two parenting dimensions predicted adolescent alcohol use (Engels et

LCOHOL IS AN IMPORTANT CONTRIBUTOR to preventable harm in young adults (Toumbourou et al., 2007). A common finding that young adult alcohol use is predicted by the level of alcohol use in adolescence and the age at initiation (Pitkänen et al., 2005; Toumbourou et al., 2004) has led to global efforts to delay the age at first alcohol use (DeWit et al., 2000; National Health and Medical Research Council, 2009). Several factors have been shown to put adolescents at risk of both early initiation and harmful alcohol use. Large longitudinal studies of adolescent alcohol use show that predictors exist in a range of areas including genetics, peer and individual characteristics, and school and family conditions (Donovan, 2004; Hawkins et al., 1992; Ryan et al., 2010). Furthermore, theories of alcohol use motivations suggest that availability and access to alcohol, together with social and physical reinforcers, are key factors influencing adolescent alcohol use (Toumbourou et al., 2007).

Received: June 24, 2014. Revision: April 8, 2015. The research was completed with funding from the National Health and Medical Research Council (Project Grant No. 251721), the Alcohol Education and Rehabilitation Foundation, and the Grosvenor Settlement Philanthropic Trust. John W. Toumbourou was supported by a VicHealth senior research fellowship. Rachel K. Leung was supported by a postdoctoral fellowship through the Deakin University Centre for Social and Early Emotional Development. Parental recruitment was supported by pro bono work from the staff of Beaton Wellmark. The funding bodies impose no contractual constraints on publishing. A potential conflict is noted in that John W. Toumbourou holds intellectual property responsibility for the Resilient Families intervention. *Correspondence may be sent to Rachel K. Leung at the School of Psychology, Faculty of Health, Deakin University, Locked Bag 20000, Geelong, Victoria 3220, Australia, or via email at: [email protected].

671

672

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / SEPTEMBER 2015

al., 2005; Ennett et al., 2001; Kosterman et al., 2000). These discrepancies may be explained by the different factors controlled in the studies. Factors shown to predict adolescent alcohol use (Hemphill et al., 2011) that were inconsistently controlled across prior parental influence studies included initial rates of alcohol and other drug use, antisocial behavior, school failure, and family context. Given that the evidence is equivocal across studies, it is important to base future analyses on theoretical models. Within the broad field of family influence, Baumrind (1991) has made a classic theoretical contribution, suggesting parenting that balanced both nurturing (e.g., communication, support) and demanding behaviors (e.g., setting and enforcing child behavior expectations) was more likely to be effective in preventing adolescent problems such as alcohol use. Baumrind’s theory regarding this balance of parental behaviors overlaps with more recent integrative theories (Catalano & Hawkins, 1996) in emphasizing the combined influence of both bonding (nurturance) and parental control practices (demands). Baumrind argued that these two parenting dimensions do not act independently but rather act interactively to form defined “parenting styles.” Baumrind defined four parenting styles, constituting different levels of nurturance and demandingness (Adalbjarnardottir & Hafsteinsson, 2001). “Authoritative” parents are those who show both nurturance (responsiveness, high bonding) and demandingness. “Authoritarian” parents are demanding and controlling but not responsive or warm. “Permissive” parents are nurturing/responsive but indulge their children by lacking demands for behavioral control. “Neglectful–Rejectful” parents are neither responsive nor demanding. Longitudinal studies show that children who were raised by authoritative parents tend to display higher levels of behavioral, emotional, and psychological adjustment (Stephenson & Helme, 2006). However, with respect to alcohol use, there has been limited research: Only four longitudinal studies have examined groups defined based on Baumrind’s parenting style typologies. In longitudinal studies (Baumrind, 1991; Lamborn et al., 1991), authoritative parenting style has been found to be protective in discouraging adolescent alcohol use. In more recent studies, neglectful parenting predicted higher rates of adolescent alcohol use (Adalbjarnardottir & Hafsteinsson, 2001; Shucksmith et al., 1997). The consistency of findings across these studies investigating parenting styles can be contrasted with the equivocal results of longitudinal studies (described above) that have examined one or two parenting behaviors independently or concurrently. Although some prior studies have investigated aspects of Baumrind’s parenting style predictions for adolescent alcohol use, the above summary suggests a number of gaps in the research to date. Hence, the aims of the present study were to test whether Baumrind’s proposed parenting styles (empirically verified by Lamborn et al., 1991) could be validly identified from Australian adolescent reports of

parenting behaviors using latent class analysis (LCA) and to explore the longitudinal prediction of the development of adolescent alcohol use comparing adolescent reports of both specific parenting behaviors and broader parenting styles. It was hypothesized that (a) four parenting styles would be identified that are comparable to the typologies proposed by Baumrind (Lamborn et al., 1991), and (b) authoritative styles would be predictive of lower levels of alcohol use relative to authoritarian, permissive-indulgent, or neglectful styles, after specific parenting behaviors and other factors were controlled for in multivariate analytic models. Method The Resilient Families Research Initiative (RFRI) is a prospective cluster randomized controlled trial evaluating the effectiveness of the Resilient Families program (Australian Clinical Trial Registry Number: 012606000399594). The Resilient Families program is an early secondary school intervention program (from Years 7 to 9) designed to develop family support networks and reduce early adolescent experience of health and social problems. Shortt and Toumbourou (2006) and Shortt and colleagues (2006) have previously detailed the program components. Measures The questionnaires used in the RFRI covered a range of factors, including items relevant to individual behaviors and attitudes and social development influences in community, family, peer, and school domains. A more detailed account of the measures can be found in Shortt et al. (2007). Specific details provided here are relevant to the items and scales used in the present analyses. Unless stated otherwise, the scales used in the present analyses were all sourced from the Communities That Care (CTC) Youth Survey (Arthur et al., 2002; Glaser et al., 2005), and scale scores consisted of the average across the included items. Parenting style groups Seventeen items were selected to represent the two key parenting dimensions that are the basis for Baumrind’s parenting styles (i.e., parental demandingness and nurturance). Parental demandingness was assessed by the CTC Low Family Management scale, which is a reliable measure incorporating items relevant to family rules and monitoring. The scale consists of six items: e.g., “My family has clear rules about alcohol and drug use”; “When I am not home, one of my parents knows where I am and who I am with” (Cronbach’s ) in the present sample = .72). Responses options were 1 = “YES!”; 2 = “yes”; 3 = “no”; and 4 = “NO!” Parental nurturance was measured using 11 items from four CTC scales. The Family Opportunities scale consists of

MINAIE ET AL. three items: for example, “My parents give me lots of chances to do fun things with them” () = .75). The Attachment to Mother scale was measured by three items: for example, “Do you feel very close to your mother?” () = .78). The Attachment to Father scale was based on the same three items as the Attachment to Mother scale with the referent changed to “father” () = .83). Responses for the above items used the response options for low Family Management, recoded such that high scores reflected higher attachment. The Family Rewards scale consists of two items: for example, “How often do your parents tell you they’re proud of you for something you’ve done?” () = .67). Response options were 1 (never or almost never), 2 (sometimes), 3 (often), and 4 (all the time). Alcohol use Alcohol use was measured at both Wave 1 and Wave 3 by asking respondents whether they had ever had more than just a sip or two of an alcoholic beverage (e.g., beer, wine, or distilled spirits). Responses were categorized into two groups: alcohol users (lifetime) and nonusers. Wave 1 predictor variables Antisocial behavior was assessed based on nine items. Four items were from the CTC survey on past antisocial behaviors asking respondents whether they had stolen something, physically attacked someone, carried a weapon, and been arrested. Response options were 0 (never), 1 (yes but not in last year), and 2 (yes, in the last year). Another five items were selected as indicators of antisocial behavior from the Strengths and Difficulties Questionnaire (Goodman, 2001): “I get very angry and often lose my temper,” “I usually do as I am told” (reverse-scored), “I fight with others,” “I am accused of lying or cheating,” and “I take things that are not mine.” Response options ranged from 0 (never) to 2 (certainly true) () = .73). Academic failure was assessed by one item: “Putting them all together, what were your marks like last year?” Response options ranged from 1 (very good) to 5 (very poor). Demographic factors were assessed based on student responses to questions on female gender, whether country of birth was outside Australia, and whether parental marital status was separated/divorced or other. Cigarette use was measured by asking respondents whether they had ever smoked cigarettes. Cannabis use was measured by asking respondents whether they had ever used marijuana. Responses for cigarette and cannabis use were categorized into two groups: users (lifetime) and nonusers. Participants and procedure The study procedures are reported in Toumbourou et al. (2013) and summarized in what follows. Thirty-nine schools from metropolitan Melbourne were randomly selected to

673

participate in the RFRI, stratified based on education sector (government or independent) and socioeconomic status (proportion of families assisted for low income). Twenty-four schools agreed to participate (62% participation rate). Before being approached, 12 schools were randomized to intervention and conducted the Resilient Families Program over 2004–2005; the remaining 12 schools were randomly assigned to be approached as usual-practice controls. For the current study, the total eligible survey population across the 24 schools consisted of students in the first year of secondary school in 2004 (Year 7; Mage = 12.3 years, SD = .50, n = 4,404). Only eligible students whose parents returned signed consent forms were approached at the baseline survey (n = 2,356, 53% of the eligible sample). Students provided their consent to participate on the day of the survey. Adolescents were resurveyed after 2 years (Year 9; Mage = 14.5 years, SD = .50). The present analysis was based on 2,081 students from Metropolitan Melbourne (47% of the eligible sample, 88% of the consenting sample; attending school in Year 7 at baseline) whose parents provided signed consent and who responded to all relevant items at the two surveys. Of the sample included in this analysis, 55.9% were female, 9.4% were born outside Australia, and 20.9% had parents who were either separated or divorced. Adolescents responded to paper questionnaires during a standard classroom period under the supervision of research staff. Students who were no longer in their original schools were followed up and surveyed in their new school locations. Analysis plan Preparation of items and variables and logistic regression models was completed using STATA Version 12 (StataCorp LP, College Station, TX); all LCA models were fit using SAS Version 9.2 for Windows (Lanza et al., 2010). LCA, being a person-centered method, was used to identify parenting style groups from the 17 items assessing parental demandingness (using items assessing Low Family Management) and nurturance (using items assessing Family Opportunities, Attachment to Mother, Attachment to Father, and Family Rewards). These items were entered into the LCA as dichotomous measures with a score of 2 reflecting agreement (e.g., collapsing “YES” “yes” on Family Management) and 1 indicating disagreement. LCA sought to identify whether item responses shared associations through a small number of discrete latent classes. Models were compared with different numbers of classes, with model selection based on (a) decreasing values in model fit statistics (e.g., the Akaike Information Criterion [AIC] and Bayesian Information Criterion [BIC]), (b) parsimony (a smaller number of classes was theoretically preferable), (c) interpretability (a solution that showed logical patterns with distinctive latent classes as revealed by the item-response probabilities), and (d) the stability and reliability of the model convergence across

674

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / SEPTEMBER 2015

1,000 trials with at least 80% of models converging to the same solution (Collins & Lanza, 2010; Finlay & Flanagan, 2013). Using the posterior probabilities of class membership generated by the final latent class model, participants were assigned into the most likely parenting style group. This “classify-analyze” approach is appropriate when the final LCA model clearly assigns membership (entropy > .80; Varvil-Weld et al., 2013). The average responses of the five CTC family scales were then calculated for each parenting style group and were used to validate and assist with labeling of the parenting style latent classes. Logistic regression analyses were performed to predict increases in adolescent alcohol use at Wave 3 after we controlled for early adolescent alcohol use at Wave 1. Alcohol use at Wave 3 was found to be significantly associated with the classroom clustering of respondents at the Wave 1 survey (intraclass correlation = .03); hence, all analyses were adjusted for classroom clustering using the STATA ‘svy’ command. The logistic regression analyses were presented in two models. In the partially adjusted models, analyses were run separately for each Wave 1 variable, while we controlled only for Wave 1 alcohol use and did not adjust for other factors. The parenting style groups were entered into these regression models as a single variable with four categories representing the four parenting styles. In the fully adjusted model, multivariate analyses were run with all Wave 1 variables entered simultaneously into the logistic regression models. Inspection of Pearson correlations revealed that the predictors included in the multivariate regression had weak to moderate associations, suggesting low collinearity. The highest correlations were between Authoritative parenting and Father Attachment (r = .65), and Authoritative parenting and Permissive parenting (r = .64). Results Latent classes of parenting styles A series of LCA models with one to six classes was run and compared to determine the optimal number of latent classes that best fit the data. The AIC and BIC suggested that the five-latent-class model was a slightly superior fit. However, the four-latent-class model (AIC = 3,767.06; BIC = 4,167.54; entropy = .81) was selected based on its parsimony, interpretability, and stable reliable convergence (100% convergence to the same solution across 1,000 trials). The five-latent-class model (AIC = 3,645.27; BIC = 4,147.29) was unreliable (35% convergence) and less interpretable. Table 1 presents the four-latent-class model groups together with the average responses to the five CTC family scales and the percentages reporting alcohol use at Wave 1 and Wave 3. Item means were interpreted as low or high within a latent class when they differed by one third (.33) or more of a standard deviation from the combined sample

mean. This difference was adequate to detect a significant effect (p < .001) in comparisons between the smallest group (n = 179) and the combined sample (N = 2,081) with more than 80% power. Class 1 was interpreted as Authoritarian (Father). Respondents reported average Family Rewards; average Mother Attachment; low Father Attachment (mean 1.91 below the combined mean of 3.16 by 1.62 times the combined standard deviation of 0.77 [i.e., effect size {ES} 1.62 = {3.16 − 1.91} / 0.77]); and average Family Opportunities and Family Management problems. Class 2 reflected Authoritative parenting based on higher than average Family rewards (ES = 0.36), Mother Attachment (ES = 0.52), Father Attachment (ES = 0.57), and Family opportunities (ES = 0.51), and low Family Management problems (ES = 0.35). Class 3 reflected a Neglectful parenting style. Although respondents reported average levels of Family rewards and Father Attachment, there were low levels of Mother Attachment (ES = 0.66), low Family Opportunities (ES = 0.48), and high Family Management problems (ES = 0.33). Class 4 was identified as a Highly Neglectful style reporting low Family Rewards (ES = 1.60), Mother Attachment (ES = 1.69), Father Attachment (ES = 1.31), and Family Opportunities (ES = 1.90), and high Family Management problems (ES = 1.39). Regression analyses predicting alcohol use Table 2 presents the logistic regression findings predicting Wave 3 alcohol use from parenting style and other factors at Wave 1. Table 2 (partially adjusted model) presents the logistic regression model predicting the prevalence of current alcohol use at Wave 3 (n = 2,081) from the parenting style groups, with Wave 1 alcohol use being adjusted for but without other predictors being adjusted for. Adolescents classified into the Highly Neglectful parenting style group were at greater risk of alcohol use in Wave 3 compared with those in the Authoritative group. The partially adjusted model showed a range of factors that were associated with Wave 3 alcohol use. The multivariate fully adjusted logistic regression model predicting current alcohol use (Wave 3) is presented in Table 2 (fully adjusted model). The result showed no significant effect of the parenting style groups. However, Poor Family Management was a significant risk factor for Wave 3 alcohol use, after other effects were adjusted for. The results revealed that alcohol use and antisocial behavior in Wave 1 increased the likelihood of Wave 3 alcohol use, whereas non-Australian birth was a protective factor. Discussion The first hypothesis was partially supported in that parenting styles were identified from adolescent reports of parenting behaviors. In line with Baumrind’s proposal (Lamborn et

MINAIE ET AL.

675

TABLE 1. Descriptive statistics for the five Communities That Care (CTC) family scales at Wave 1 and percentage reporting alcohol use in Waves 1 and 3 of each latent class group Latent class 1 Authoritarian (n = 227, 10.9%)

Interpretation (size) Wave 1 CTC family scales Family Rewards Mother Attachment Father Attachment Family Opportunities Poor Family Management Alcohol use Wave 1 Alcohol use Wave 3 Alcohol use

2 Authoritative (n = 1,186, 57.0%)

3 Neglectful (n = 489, 23.5%)

4 Highly Neglectful (n = 179, 8.6%)

Combined sample (N = 2,081)

M

SD

M

SD

M

SD

M

SD

M

SD

2.96 3.45 1.91 3.16 1.51

0.78 0.50 0.60 0.55 0.42

3.42 3.74 3.60 3.59 1.30

0.63 0.32 0.41 0.41 0.32

2.98 3.02 3.06 2.91 1.61

0.69 0.53 0.45 0.57 0.45

1.85 2.39 2.15 1.93 2.10

0.77 0.64 0.71 0.57 0.58

3.13 3.42 3.16 3.24 1.46

0.80 0.61 0.77 0.69 0.46

%

[95% CI]

%

[95% CI]

%

[95% CI]

%

[95% CI]

%

[95% CI]

34 73

[27, 41] [68, 78]

27 69

[24, 30] [66, 72]

40 77

[35, 45] [73, 81]

51 82

[44, 59] [76, 88]

33 72

[30, 35] [70, 75]

Note: CI = confidence interval.

al. 1991), a four-latent-class model was found to be the best fit to the data. However, the four-class model differed from the Baumrind typology in not identifying a permissive–indulgent parenting style group. The second hypothesis, that authoritative parenting would predict lower rates of alcohol use in adolescents after specific parenting behavior domains and other factors were controlled for, was not supported. The findings demonstrated that parenting styles did influence adolescent alcohol use in the partially adjusted model, where the domains of parenting behaviors and other predictors were not included simultaneously in the regression model. However, when the regression model was fully adjusted for the direct effects of family management, parental nurturance, TABLE 2.

and other predictors, the effect of parenting style was no longer significant. The finding that parenting style groups were able to be modeled from adolescent reports of parenting behaviors supported Chan and Koo’s (2011) results and demonstrated that parenting styles were identifiable within a large Australian sample. In the present analysis, the four-class model differed from the Baumrind typology in not identifying a group with high parental nurturance but low demandingness. This finding was similar to that of Chan and Koo (2011), who identified only three latent groups in their analysis of adolescents in the United Kingdom. In recent decades, maternal workforce participation has increased, with the consequence that

Logistic regression predicting Wave 3 alcohol use from variables measured in Wave 1 Partially adjusted modela

Wave 1 predictors

OR

Parenting style (referent: Authoritative) Authoritarian Neglectful Highly Neglectful Family Rewards Mother Attachment Father Attachment Family Opportunities Poor Family Management Female Non-Australian Birth Parents Separated/Divorced Alcohol Use Cannabis Use Cigarette Use Academic Failure Antisocial Behavior Intervention

1.13 1.29 1.56* 0.90 0.85 0.84* 0.82* 1.80** 1.08 0.56** 1.33* 4.46** 2.14 1.96* 1.20* 3.47** 0.83

[95% CI]

[0.84, 1.51] [0.99, 1.68] [1.04, 2.35] [0.80, 1.02] [0.71, 1.02] [0.74, 0.96] [0.71, 0.95] [1.40, 2.33] [0.88, 1.32] [0.41, 0.78] [1.01, 1.73] [3.42, 5.83] [0.27, 17.16] [1.06, 3.64] [1.04, 1.39] [2.21, 5.43] [0.66, 1.03]

Fully adjusted modelb OR

0.73 1.02 0.86 1.04 1.11 0.87 1.00 1.56* 1.19 0.57** 1.18 3.19** 0.99 1.39 1.09 3.07** 0.81

[95% CI]

[0.45, 1.21] [0.72, 1.44] [0.44, 1.69] [0.88, 1.23] [0.84, 1.48] [0.67, 1.12] [0.80, 1.26] [1.16, 2.10] [0.96, 1.47] [0.41, 0.79] [0.88, 1.57] [2.38, 4.29] [0.16, 6.04] [0.72, 2.69] [0.93, 1.27] [1.88, 5.03] [0.65, 1.02]

Notes: OR = odds ratio; CI = confidence interval. Bold indicates statistical significance. aAnalyses were run separately for each predictor controlling only for Wave 1 alcohol use and excluding the other factors in the table. The Parenting Style Groups identified from the latent class analysis were entered as a group and compared with the Authoritative group; banalyses controlled for all predictors. *p < .05; **p $ .001.

676

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / SEPTEMBER 2015

some Australian families now experience work–family conflicts that can increase the potential for child neglect (Strazdins et al., 2013) and may reduce the possibility of indulging children. The Authoritarian and Neglectful groups identified in the present analysis were characterized by divergences in evaluations of mother and father attachments. These results may reflect the fact that rates of family breakdown are now higher relative to Baumrind’s (1991) context. Contrary to the hypothesis based on Baumrind’s theory, parenting style was found to have no significant effect in predicting adolescent alcohol use after the direct effects of other factors were adjusted for. This result is consistent with the Engels et al. (2005) finding that the interaction effect between affection and control (measuring low and high combinations of these variables) was not significant in predicting adolescent progression to heavy alcohol use. The present findings suggest that the overall parenting style (i.e., the distinct balance of parental responsiveness and demandingness) may be less important than specific family management practices (i.e., setting clear rules and effectively monitoring behavior) in predicting adolescent alcohol use. The current study demonstrated that parental nurturance measures were predictive of adolescent alcohol use in the partially adjusted models but not significant after the effect of family management and other factors was adjusted for. This is consistent with the majority of studies that have examined the effect of both parenting domains and found that only parental monitoring was independently predictive of adolescent alcohol use (Aquilino & Supple, 2001; Getz & Bray, 2005; Huh et al., 2006; Webb et al., 2002). Some studies have found that parental warmth/nurture independently predicted adolescent alcohol use in analyses that controlled for family monitoring. Comparable studies differed from the present study in controlling for fewer factors. For instance, none of the variables controlled for in the present study was controlled for in Nash et al. (2005), Pires and Jenkins (2007), and Van Der Zwaluw et al. (2008). Brody and Ge (2001) and Brody et al. (2009) measured both parenting domains with a single variable. Latendresse et al. (2008) adjusted fewer factors than the present study. In contrast with the majority of studies and the present findings, three prior studies found no effect for family management (Engels et al., 2005; Ennett et al., 2001; Kosterman et al., 2000). In the Ennett et al. (2001) study, the measure of parental monitoring had a relatively low reliability () = .50). Engels et al. (2005) found no effect for family management measured as child-perceived control, but family functioning measured as child and parental ratings of order and structure was significant in the adjusted analyses. It was likely that this variable captured aspects of the family environment that underlie family management. In Kosterman et al. (2000), family management was shown to be significant, but not after parental alcohol use was adjusted for. The present findings in predicting adolescent alcohol use

are in line with several studies (e.g., Aquilino & Supple, 2001) that found parental nurture domains were significant before full adjustment of other factors, but not after. These findings may be consistent with Barnes et al.’s (2000) proposition that parental support is indirectly related to adolescent alcohol use through monitoring. In common with previous studies (Adalbjarnardottir & Hafsteinsson, 2001; Dishion et al., 1999; Shucksmith et al., 1997), the present study found that Highly Neglectful parenting increases the likelihood of adolescent alcohol use relative to Authoritative parenting. The present findings showed only a marginally significant difference between Authoritative and Neglectful parenting styles as predictors of adolescent alcohol use (p = .055). However, Table 1 suggested differences in Year 7 rates of alcohol use between Authoritative and Neglectful parenting styles. That a significant difference was not found may be due to the current sample size being slightly underpowered. The present finding that Authoritative and Authoritarian parenting predicted a similar longitudinal risk for the development of adolescent alcohol use was unexpected because it contradicted previous longitudinal studies (Adalbjarnardottir & Hafsteinsson, 2001; Shucksmith et al., 1997). However, the present study was distinctive in controlling for Wave 1 alcohol use. A key predictor in the present study was alcohol use in Wave 1. Previous studies have found that early adolescent alcohol use is a major predictor for later use in adolescence (e.g., Webb et al., 2002). One of the strengths of the present study was that the regression analyses controlled for the effects of early adolescent alcohol use. Prior longitudinal studies suggest that parental behaviors may have less influence on adolescent alcohol use after adolescents have initiated alcohol use (Huh et al., 2006; Van Der Zwaluw et al., 2008). Consistent with Engels et al. (2005) and Getz and Bray (2005), antisocial behavior was found to increase the risk of adolescent alcohol use in both models. Although Huh et al. (2006) did not directly examine whether externalizing behavior predicts later alcohol use, they did demonstrate that externalizing behavior influences parenting. The present findings confirm that child antisocial behavior and family management (parental monitoring) both independently predict adolescent alcohol use. The partially adjusted results support the view that bonding with parents is important in predicting adolescent alcohol use. However, this effect was no longer apparent once family management and other factors were included in the fully adjusted model. Although mediation analyses were not completed, the findings are compatible with Barnes et al. (2000), who found that family bonding acted indirectly by mediating parental monitoring practices. Consistent with the proposition that parental bonds may be most important at earlier ages, Van Der Zwaluw et al. (2008) found that a direct effect of parenting nurturing behaviors was only evident in younger adolescents.

MINAIE ET AL. Limitations Before we consider the implications of the present study, we must consider potential limitations. First, it is important to recognize that although this study was longitudinal, the test for parenting effects on adolescent alcohol use was based on a nonexperimental design. Specifically, the predictor variables were not subject to manipulation through intervention. Thus, the study cannot strongly claim to have identified causal effects. Second, the classification of parenting style was based on a specific selection of items and adolescent reports of parenting. It is possible that other methods of classifying parenting style may have produced different results. For example, it may be important to look not just at adolescent reports, but also at the parents’ reports of parenting style. Third, the analytic approach used logistic regression to identify predictors from the identified latent class groups. This “classify–analyze” approach has been criticized because it does not take into account the uncertainty of class assignment and can lead to biased prediction estimates. In support of the present analysis, the final models clearly assigned membership (entropy > .80), reducing bias in the prediction analyses (Bray et al., 2012). The fourth limitation is that the present analyses did not focus on the potential for parenting style to mediate and moderate the effects of parenting behavior (Darling & Steinberg, 1993; Spera, 2005). Exploratory investigations in the present sample (not reported) did not find indications of parental style moderation in the prediction of Wave 3 alcohol use. Fifth, the present study did not control for some of the variables found to be important in previous studies (e.g., parental alcohol use; Kosterman et al., 2000). A number of scales had a relatively low reliability (Cronbach’s )), and this may underestimate their impact on adolescent alcohol use. Last, it should be noted that the present results cannot be generalized beyond the recruited sample within the metropolitan schools included in the current study. A relatively small proportion of the families within the recruited schools returned signed consent forms. Despite this problem, the participating students in this study reported similar rates of alcohol use compared with those in statewide samples (Toumbourou et al., 2013). Future research should continue to study parenting style in a range of populations. Implications for research and health-promotion practice In common with previous research, the current study shows that early adolescent alcohol use is a major predictor of later adolescent alcohol use. This finding confirms the importance of health-promotion programs that seek to delay or prevent adolescent alcohol use. The present findings noted a trend for random assignment to the Resilient Families intervention to be associated with reduced levels of adolescent alcohol use. Although not reported here, the intervention was

677

found to significantly reduce the amount and frequency of alcohol use (Toumbourou et al., 2013). One effective component of the intervention was the encouragement given to parents to set household rules that did not supply or allow adolescent alcohol use (Toumbourou et al., 2013). The current findings emphasize family management practices that include establishing clear rules for alcohol and other drugs and effective monitoring practices as important actions that parents can adopt in efforts to reduce adolescent alcohol use. Although parenting style did not directly affect adolescent alcohol use during early secondary school, Table 1 revealed that children in authoritative families entered secondary school (Year 7, Wave 1) with lower rates of alcohol use. This finding opens the possibility that there may have been benefits in earlier developmental periods. It was argued previously by Toumbourou et al. (2013) that secondary school parenting interventions targeting adolescent alcohol use may be usefully advised to focus on family management practices. It is plausible that the influence of parenting style may have more specific importance for preventing adolescent alcohol use in late childhood. Therefore, it may be advisable to target family intervention/prevention programs for alcohol use in primary schools in countries such as Australia that have high rates of early adolescent alcohol use (Toumbourou et al., 2014). Conclusions The present study found that adolescents could provide reliable (internally consistent) descriptions of parenting behaviors and these behaviors were used in LCAs to identify four parenting style groups that had some similarities and differences when compared with the typologies proposed by Baumrind. The hypothesis that authoritative parenting would predict lower rates of alcohol use in adolescents was not fully supported. Rather, family management was a more direct predictor of adolescent alcohol use than was parenting style at Wave 3 in the multivariate analysis (Table 2, Fully Adjusted Model). These findings highlight the need to critically evaluate the role of parenting style in prior research and to conduct further research across different developmental periods, using a wider range of measures, methods, and theoretically guided approaches. References Adalbjarnardottir, S., & Hafsteinsson, L. G. (2001). Adolescents’ perceived parenting styles and their substance use: Concurrent and longitudinal analyses. Journal of Research on Adolescence, 11, 401–423. doi:10.1111/1532-7795.00018. Aquilino, W. S., & Supple, A. J. (2001). Long-term effects of parenting practices during adolescence on well-being outcomes in young adulthood. Journal of Family Issues, 22, 289–308. doi:10.1177/019251301022003002. Arthur, M. W., Hawkins, J. D., Pollard, J. A., Catalano, R. F., & Baglioni, A.

678

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / SEPTEMBER 2015

J., Jr. (2002). Measuring risk and protective factors for substance use, delinquency, and other adolescent problem behaviors. The Communities That Care Youth Survey. Evaluation Review, 26, 575–601. Bahr, S. J., & Hoffmann, J. P. (2010). Parenting style, religiosity, peers, and adolescent heavy drinking. Journal of Studies on Alcohol and Drugs, 71, 539–543. doi:10.15288/jsad.2010.71.539. Barnes, G. M., Reifman, A. S., Farrell, M. P., & Dintcheff, B. A. (2000). The effects of parenting on the development of adolescent alcohol misuse: A six-wave latent growth model. Journal of Marriage and Family, 62, 175–186. doi:10.1111/j.1741-3737.2000.00175.x. Baumrind, D. (1991). The influence of parenting style on adolescent competence and substance use. Journal of Early Adolescence, 11, 56–95. doi:10.1177/0272431691111004. Bray, B., Lanza, S. T., & Tan, X. (2012). An introduction to eliminating bias in classify-analyze approaches for latent class analysis. University Park, PA: The Methodology Center at Penn State. Brody, G. H., Beach, S. R. H., Philibert, R. A., Chen, Y. F., Lei, M. K., Murry, V. M., & Brown, A. C. (2009). Parenting moderates a genetic vulnerability factor in longitudinal increases in youths’ substance use. Journal of Consulting and Clinical Psychology, 77, 1–11. doi:10.1037/ a0012996. Brody, G. H., & Ge, X. (2001). Linking parenting processes and self-regulation to psychological functioning and alcohol use during early adolescence. Journal of Family Psychology, 15, 82–94. doi:10.1037/0893-3200.15.1.82. Brook, J. S., Brook, D. W., Gordon, A. S., Whiteman, M., & Cohen, P. (1990). The psychosocial etiology of adolescent drug use: A family interactional approach. Genetic, Social, and General Psychology Monographs, 116, 111–267. Catalano, R. F., & Hawkins, J. D. (1996). The Social Development Model: A theory of antisocial behavior. In J. D. Hawkins (Ed.), Delinquency & crime: Current theories. New York, NY: Cambridge University Press. Chan, T. W., & Koo, A. (2011). Parenting style and youth outcomes in the UK. European Sociological Review, 27, 385–399. doi:10.1093/esr/ jcq013. Clausen, S.-E. (1996). Parenting styles and adolescent drug use behaviors. Childhood, 3, 403–414. doi:10.1177/0907568296003003006. Coker, J. K., & Borders, L. D. (2001). An analysis of environmental and social factors affecting adolescent problem drinking. Journal of Counseling & Development, 79, 200–208. doi:10.1002/j.1556-6676.2001. tb01961.x. Collins, L. M., & Lanza, S. T. (2010). Latent class and latent transition analysis: With applications in the social, behavioral, and health sciences. Hoboken, NJ: John Wiley & Sons. Darling, N., & Steinberg, L. (1993). Parenting style as context: An integrative model. Psychological Bulletin, 113, 487–496. doi:10.1037/0033-2909.113.3.487. 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. doi:10.1176/appi. ajp.157.5.745. Dishion, T. J., Capaldi, D. M., & Yoerger, K. (1999). Middle childhood antecedents to progressions in male adolescent substance use: An ecological analysis of risk and protection. Journal of Adolescent Research, 14, 175–205. doi:10.1177/0743558499142003. Donovan, J. E. (2004). Adolescent alcohol initiation: A review of psychosocial risk factors. Journal of Adolescent Health, 35, 529.e7–529.e18. doi:10.1016/S1054-139X(04)00066-7. Engels, R. C., Vermulst, A. A., Dubas, J. S., Bot, S. M., & Gerris, J. (2005). Long-term effects of family functioning and child characteristics on problem drinking in young adulthood. European Addiction Research, 11, 32–37. doi:10.1159/000081414. Ennett, S. T., Bauman, K. E., Foshee, V. A., Pemberton, M., & Hicks, K. A. (2001). Parent-child communication about adolescent to-

bacco and alcohol use: What do parents say and does it affect youth behavior? Journal of Marriage and Family, 63, 48–62. doi:10.1111/j.1741-3737.2001.00048.x. Finlay, A. K., & Flanagan, C. (2013). Adolescents’ civic engagement and alcohol use: Longitudinal evidence for patterns of engagement and use in the adult lives of a British cohort. Journal of Adolescence, 36, 435–446. doi:10.1016/j.adolescence.2013.01.006. Getz, J. G., & Bray, J. H. (2005). Predicting heavy alcohol use among adolescents. American Journal of Orthopsychiatry, 75, 102–116. doi:10.1037/0002-9432.75.1.102. Glaser, R. R., Van Horn, M. L., Arthur, M. W., Hawkins, J. D., & Catalano, R. F. (2005). Measurement properties of the Communities That Care youth survey across demographic groups. Journal of Quantitative Criminology, 21, 73–102. doi:10.1007/s10940-004-1788-1. Goodman, R. (2001). Psychometric properties of the strengths and difficulties questionnaire. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337–1345. doi:10.1097/00004583-200111000-00015. 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. doi:10.1037/0033-2909.112.1.64. Hemphill, S. A., Heerde, J. A., Herrenkohl, T. I., Patton, G. C., Toumbourou, J. W., & Catalano, R. F. (2011). Risk and protective factors for adolescent substance use in Washington state, the United States and Victoria, Australia: A longitudinal study. Journal of Adolescent Health, 49, 312–320. doi:10.1016/j.jadohealth.2010.12.017. Huh, D., Tristan, J., Wade, E., & Stice, E. (2006). Does problem behavior elicit poor parenting? A prospective study of adolescent girls. Journal of Adolescent Research, 21, 185–204. doi:10.1177/0743558405285462. Koning, I. M., van den Eijnden, R. J., Glatz, T., & Vollebergh, W. A. (2013). Don’t worry! Parental worries, alcohol-specific parenting and adolescents’ drinking. Cognitive Therapy and Research, 37, 1079–1088. doi:10.1007/s10608-013-9545-0. Koning, I. M., van den Eijnden, R. J. J. M., Verdurmen, J. E. E., Engels, R. C. M. E., & Vollebergh, W. A. M. (2012). Developmental alcoholspecific parenting profiles in adolescence and their relationships with adolescents’ alcohol use. Journal of Youth and Adolescence, 41, 1502–1511. doi:10.1007/s10964-012-9772-9. Kosterman, R., Hawkins, J., Guo, J., Catalano, R., & Abbott, R. (2000). The dynamics of alcohol and marijuana initiation: Patterns and predictors of first use in adolescence. American Journal of Public Health, 90, 360–366. Lamborn, S. D., Mounts, N. S., Steinberg, L., & Dornbusch, S. M. (1991). Patterns of competence and adjustment among adolescents from authoritative, authoritarian, indulgent, and neglectful families. Child Development, 62, 1049–1065. doi:10.2307/1131151. Lanza, S. T., Lemmon, D., Dziak, J. J., Huang, L., Schafer, J. L., & Collins, L. M. (2010). PROC LCA & PROC LTA User’s Guide, Version 1.2.5. University Park, PA: The Methodology Center, Penn State. Latendresse, S. J., Rose, R. J., Viken, R. J., Pulkkinen, L., Kaprio, J., & Dick, D. M. (2008). Parenting mechanisms in links between parents’ and adolescents’ alcohol use behaviors. Alcoholism: Clinical and Experimental Research, 32, 322–330. doi:10.1111/j.1530-0277.2007.00583.x. Mares, S. H., Lichtwarck-Aschoff, A., Burk, W. J., van der Vorst, H., & Engels, R. C. (2012). Parental alcohol-specific rules and alcohol use from early adolescence to young adulthood. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 53, 798–805. doi:10.1111/j.1469-7610.2012.02533.x. Nash, S. G., McQueen, A., & Bray, J. H. (2005). Pathways to adolescent alcohol use: Family environment, peer influence, and parental expectations. Journal of Adolescent Health, 37, 19–28. doi:10.1016/j. jadohealth.2004.06.004. National Health and Medical Research Council. (2009). Australian guide-

MINAIE ET AL. lines to reduce health risks from drinking alcohol. Commonwealth of Australia. Pires, P., & Jenkins, J. M. (2007). A growth curve analysis of the joint influences of parenting affect, child characteristics and deviant peers on adolescent illicit drug use. Journal of Youth and Adolescence, 36, 169–183. doi:10.1007/s10964-006-9127-5. Pitkänen, T., Lyyra, A.-L., & Pulkkinen, L. (2005). Age of onset of drinking and the use of alcohol in adulthood: A follow-up study from age 8-42 for females and males. Addiction, 100, 652–661. doi:10.1111/j.1360-0443.2005.01053.x. Ryan, S. M., Jorm, A. F., & Lubman, D. I. (2010). Parenting factors associated with reduced adolescent alcohol use: A systematic review of longitudinal studies. Australian and New Zealand Journal of Psychiatry, 44, 774–783. doi:10.1080/00048674.2010.501759. Shortt, A. L., Hutchinson, D. M., Chapman, R., & Toumbourou, J. W. (2007). Family, school, peer and individual influences on early adolescent alcohol use: First-year impact of the Resilient Families programme. Drug and Alcohol Review, 26, 625–634. doi:10.1080/09595230701613817. Shortt, A., & Toumbourou, J. W. (2006). The Resilient Families Program: Helping to prepare adolescents for success in school and life. Youth Studies Australia, 25(1), 57–58. Shortt, A., Toumbourou, J. W., Chapman, R., & Power, E. (2006). The Resilient Families Program: Promoting health and wellbeing in adolescents and their parents during the transition to secondary school. Youth Studies Australia, 25(2), 33–40. Shucksmith, J., Glendinning, A., & Hendry, L. (1997). Adolescent drinking behaviour and the role of family life: A Scottish perspective. Journal of Adolescence, 20, 85–101. doi:10.1006/jado.1996.0066. Spera, C. (2005). A review of the relationship among parenting practices, parenting styles, and adolescent school achievement. Educational Psychology Review, 17, 125–146. doi:10.1007/s10648-005-3950-1. Stephenson, M. T., & Helme, D. W. (2006). Authoritative parenting and sensation seeking as predictors of adolescent cigarette and marijuana use. Journal of Drug Education, 36, 247–270.

679

Stice, E., Barrera, M., Jr., & Chassin, L. (1998). Prospective differential prediction of adolescent alcohol use and problem use: Examining the mechanisms of effect. Journal of Abnormal Psychology, 107, 616–628. doi:10.1037/0021-843X.107.4.616. Strazdins, L., O’Brien, L. V., Lucas, N., & Rodgers, B. (2013). Combining work and family: Rewards or risks for children’s mental health? Social Science & Medicine, 87, 99–107. doi:10.1016/j.socscimed.2013.03.030. Toumbourou, J. W., Gregg, M. E., Shortt, A. L., Hutchinson, D. M., & Slaviero, T. M. (2013). Reduction of adolescent alcohol use through family-school intervention: A randomized trial. Journal of Adolescent Health, 53, 778–784. doi:10.1016/j.jadohealth.2013.07.005. Toumbourou, J. W., Olsson, C. A., Rowland, B., Renati, S., & Hallam, B. (2014). Health psychology intervention in key social environments to promote adolescent health. Australian Psychologist, 49, 66–74. doi:10.1111/ap.12043. Toumbourou, J. W., Stockwell, T., Neighbors, C., Marlatt, G. A., Sturge, J., & Rehm, J. (2007). Interventions to reduce harm associated with adolescent substance use. The Lancet, 369, 1391–1401. doi:10.1016/ S0140-6736(07)60369-9. Toumbourou, J. W., Williams, I. R., White, V. M., Snow, P. C., Munro, G. D., & Schofield, P. E. (2004). Prediction of alcohol-related harm from controlled drinking strategies and alcohol consumption trajectories. Addiction, 99, 498–508. doi:10.1111/j.1360-0443.2004.00689.x. van der Zwaluw, C. S., Scholte, R. H. J., Vermulst, A. A., Buitelaar, J. K., Verkes, R. J., & Engels, R. C. (2008). Parental problem drinking, parenting, and adolescent alcohol use. Journal of Behavioral Medicine, 31, 189–200. doi:10.1007/s10865-007-9146-z. Varvil-Weld, L., Marzell, M., Turrisi, R., Mallett, K. A., & Cleveland, M. J. (2013). Examining the relationship between alcohol-energy drink risk profiles and high-risk drinking behaviors. Alcoholism: Clinical and Experimental Research, 37, 1410–1416. doi:10.1111/acer.12102. Webb, J. A., Bray, J. H., Getz, J. G., & Adams, G. (2002). Gender, perceived parental monitoring, and behavioral adjustment: Influences on adolescent alcohol use. American Journal of Orthopsychiatry, 72, 392–400. doi:10.1037/0002-9432.72.3.392.