Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jan 1.
Published in final edited form as: J Adolesc. 2019 Dec 5;78:24–32. doi: 10.1016/j.adolescence.2019.11.007

Anhedonia, Screen Time, and Substance Use in Early Adolescents: A Longitudinal Mediation Analysis

Georgia Christodoulou 1, Anuja Majmundar 1, Chih-Ping Chou 1, Mary Ann Pentz 1
PMCID: PMC6935414  NIHMSID: NIHMS1546514  PMID: 31812941

Abstract

Introduction

Screen time is associated with substance use risk among adolescents; however, less is known about the underlying factors that explain this association.

Methods

This research examined anhedonia, a facet of depression noted by the reduced capacity to experience pleasure, as a mediating factor in the relationship between screen time (hours spent watching television, using internet, and/or playing video games outside of school) and substance use (alcohol and/or cigarettes), using a longitudinal survey design among a sample of students aged 9–11 years in the 4th to 6th grades in Southern California, United States of America [N=709 (354 males)].

Results

Structural equation modeling (SEM) findings revealed longitudinal mediation effects of anhedonia on the association between screen time and substance use, while controlling for baseline screen time, baseline individual and peer substance use, race/ethnicity, and gender. Moderation analysis based on a multiple-group approach revealed that gender was not a significant moderator of this mediation pathway.

Conclusions

Adolescents may become desensitized and exhibit a blunted response to hedonic effects from increased screen time. This may result in increased anhedonia and greater risk for substance use through the need to compensate for the reduced experience of rewards. These findings have implications for future school-based substance use prevention and intervention programs.

Keywords: Anhedonia, Screen Time, Substance Use, Adolescents, Alcohol, Cigarettes

Introduction

Adolescence typically refers to the ages between 10 and 19 years old according to the World Health Organization (“Age limits and adolescents,” 2003) and represents a period of major developmental upheaval and growth. Deficits in self-regulatory capacity in adolescence and difficulties in impulse control increase adolescents’ susceptibility to maladaptive behaviors, such as substance use (Andersen & Teicher, 2008; Bechara, 2005; Casey, Jones, & Hare, 2008). Substance use in early adolescence has been associated with a plethora of adverse health outcomes and behaviors in adulthood, including the development of substance use disorders, criminal behavior, risky sexual behavior, unemployment, and difficulties in interpersonal relationships (Nelson, van Ryzin, & Dishion, 2015; Stueve & O’Donnell, 2005; Welty et al., 2017). Therefore, targeting adolescents earlier on in this developmental trajectory becomes an integral component to substance use prevention efforts, particularly in light of heightened risks associated with early substance use.

In addition to substance use risk, younger adolescents spend an increasing amount of time in front of screens with current studies reporting averages of up to 3 hours per day (Twenge & Campbell, 2018). Screen time has historically been assessed as an indicator of sedentary behavior (primarily television and computer use)(Saunders & Vallance, 2017; Sisson et al., 2009), and investigations have tended to focus on associations with metabolic syndrome and obesity (Saunders & Vallance, 2017). In recent years, researchers have expanded their studies on screen time to focus more on behavioral and mental health outcomes, including substance use and depression (Ha et al., 2007; Hanewinkel & Sargent, 2009; Hull, Brunelle, Prescott, & Sargent, 2014; Ko et al., 2008; Maras et al., 2015; Park, Hong, Park, Ha, & Yoo, 2013; Primack, Kraemer, Fine, & Dalton, 2009; Rucker, Akre, Berchtold, & Suris, 2015).

Studies suggest that increased screen time (watching television, using the internet, and playing video games) and/or exposure to harmful media content can potentially trigger and exacerbate negative health behaviors, including experimentation with illegal substances (Carson, Pickett, & Janssen, 2011; Hanewinkel & Sargent, 2009; Hull et al., 2014; Ko et al., 2008; Rucker et al., 2015). In a cross-sectional study of Canadian youth, increased screen time (particularly computer and television use) was associated with a composite multiple risk behavior score, comprised of six variables that included lifetime drunkenness and cigarette use (Carson et al., 2011). In a longitudinal study that also assessed risk behaviors, researchers found that engagement with risk-glorifying video games led to increased alcohol use, cigarette smoking, aggression, delinquency, and risky sexual behavior, respectively (Hull et al., 2014). A cross-sectional study focusing on content-based media exposure found that alcohol use exposure in movies was associated with binge drinking across six European countries. Cross-sectional studies that assessed problematic internet use and addiction among middle and high school students have also reported associations with substance use, specifically alcohol and cigarettes (Ko et al., 2008; Rucker et al., 2015). The majority of these studies used cross-sectional versus longitudinal designs; therefore, linear associations between screen time and substance use can only be inferred. Additionally, studies differed in terms of focusing on screen content versus dosage (hours spent in front of screens). Therefore, it is still unclear what specifically drives the association between screen time and increased risk of substance use in early adolescence.

In addition to maladaptive behavioral outcomes such as substance use, recent studies have begun to examine the association between screen time and mental health, specifically depression and related symptomology. A meta-analysis that assessed twelve cross-sectional and four longitudinal studies among adolescents showed a significant dose-response association between screen time and risk of depression (Liu, Wu, & Yao, 2015). Moreover, a systematic review on sedentary behavior among 32 studies found strong associations between amount of time spent in front of a screen and depressive symptoms (Hoare, Milton, Foster, & Allender, 2016). Maras and colleagues (2015) found that increased screen time (specifically computer and video game playing) was associated with symptoms of depression among a sample of Canadian youth. A very recent cross-sectional study reported significant associations between increased screen time and depressive symptoms and suicides, particularly among girls (Twenge, Joiner, Rogers, & Martin, 2018). The majority of these studies all point to a dose-response relationship between hours spent in front of a screen and risk of depression across adolescence.

Despite the associations between screen time and depression and screen time and substance use, longitudinal investigations that assess how both these constructs interact to drive substance use have not been undertaken in the current literature. Taking into consideration the increasing amount of time adolescents spend in front of screens, identification of target indicators for substance use prevention efforts and interventions among early adolescents is needed. Nuanced assessment of potential factors that may underly the associations between screen time, depression, and substance use could help identify and target at-risk youth.

Anhedonia may represent an important transdiagnostic indicator that can provide insight on the association between screen time and substance use (Cuthbert & Insel, 2013; Leventhal & Zvolensky, 2015). Anhedonia is defined as the reduced capacity of experiencing pleasure or positive emotions and potentially manifests as a lack of motivation or drive to pursue enjoyable activities (Husain & Roiser, 2018). Although typically characterized as a symptom of depression, anhedonia is not highly correlated with other depressive symptoms and can be prevalent across many other psychiatric conditions (Leventhal & Zvolensky, 2015). Due to these unique characteristics, anhedonia is considered a transdiagnostic indicator of emotion vulnerability underlying multiple psychopathologies (depression, anxiety, etc.) and may represent a distinctive, maladaptive indicator for substance use risk (Carleton et al., 2013; Leventhal & Zvolensky, 2015). Moreover, anhedonia can be a relatively stable characteristic seen in healthy adolescent populations as well (Sussman, & Leventhal, 2014). Although the assessment of anhedonia among early adolescents is still limited (Leventhal et al. 2015) and important data concerning prevalence rates are scarce, emerging evidence from recent studies has linked anhedonia with substance use among youth, including both cigarette smoking and alcohol (Cano et al., 2017; Chuang, Chan, & Leventhal, 2016; Leventhal & Zvolensky, 2015; Leventhal et al., 2016).

Within the past few years, researchers have proposed explanations for the association of anhedonia with substance use among adolescents. Sussman and Leventhal (2014) speculate that anhedonic adolescents seek out hedonic-driven behavior, such as substance use, in an effort to compensate for deficits in the experience of rewards. In other words, anhedonic adolescents may find activities that are characterized by low to moderate intensity unrewarding. In turn, these adolescents may be drawn to substance use, given the relatively intense pharmacological effects of illegal substances on the brain’s dopamine reward pathway (Sussman & Leventhal, 2014). Although there have been numerous recent studies that have assessed the effects of increased screen time on depression (Maras et al., 2015; Boers, Afzali, Newton, & Conrod, 2019; Hoare et al. 2016, Twenge et al. 2018), no studies to our knowledge have assessed the possible linear association between screen time and anhedonia among early adolescents. Given the prevalence of anhedonia among youth (Gunzler, Chen, Wu, & Zhang, 2013; Leventhal et al., 2015; McMakin et al., 2012), investigating the mediating role of anhedonia between screen time and substance use is warranted, and findings may point to unique target indicators for substance use prevention efforts.

Gender may moderate the relationship between screen time, anhedonia, and substance use, particularly due to gender differences among these constructs of interest. While males tend to show greater substance use behaviors in mid to late adolescence, girls tend to report similar or higher rates of substance use in early adolescence (Chen & Jacobson, 2012; Kuhn, 2015). Prevalence of depression among adolescent girls is notably higher than boys (Saluja et al., 2004). Despite this finding, anhedonia may represent an independent transdiagnostic indicator that is expressed differently between genders. Studies have reported higher prevalence of anhedonia among boys compared to girls in early and middle adolescence (Bennett, Ambrosini, Kudes, Metz, & Rabinovich, 2005; Khesht-Masjedi, Shokrgozar, Abdollahi, Golshahi, & Sharif-Ghaziani, 2017). There is no general consensus on whether one gender engages in more screen time than the other (Hands et al., 2008; Foltz et al., 2011). Despite mixed results across studies, boys generally engage in video game playing much more frequently than girls and thus may be spending comparatively more time exposed to content that reinforces risk-taking behaviors (Hull et al., 2014; Van Rooij et al. 2014; Smith, Hummer, & Hulvershorn, 2015). Effects of screen time on anhedonia and substance use may be more prominent among boys, and early adolescent boys who engage in more screen time may represent a particularly at-risk population. Thus, investigating the interrelationships of screen time, anhedonia, and substance use and assessing the effects of gender may guide the development of more effective interventions (Pang, Farrahi, Glazier, Sussman, & Leventhal, 2014; Saluja et al., 2004; Waller et al., 2006).

The current study sought to investigate the underlying pathways between anhedonia, screen time, and substance use among a sample of racially and ethnically diverse early adolescents. We hypothesized that anhedonia at follow-up would mediate the relationship between screen time and substance use among youth. We predicted that the mediation effect of anhedonia on the association between screen time and substance use would be stronger among boys.

Methods

Sample

The sample for this study was derived from the 4th to 6th grade Pathways to Health randomized control trial (RCT) involving a sample of 1005 4th grade students (aged 9 – 11 years) from 28 elementary schools in Southern California in the United States of America. Students were randomized by school into either a prevention program or control condition and followed for 3 years from 2009 to 2011 (Pentz & Riggs, 2013). The purpose of this parent RCT was to evaluate the effects of a substance and obesity prevention program among elementary school children (Sakuma, Riggs, & Pentz, 2012). Enrolled students with active parental consent and self-assent were assessed via multiple measures across 4 waves of data collection (Wave A-D) from 2009–2011. Demographics of the sample have been previously reported, including an examination of differences between participants with complete data (N=709) and those with missing data; the results showed that the analysis sample was representative of the school populations from which the sample was drawn (Pentz & Riggs, 2013; Shin et al., 2014). Fifty-two cases (7% of the data) had missing entries. A closer look at the differences between cases with missing and non-missing data revealed no significant differences between groups on demographic variables or variables of interest, including anhedonia, screen time, and substance use. The present study utilized longitudinal data from self-report surveys administered during the spring semesters of 4th grade (baseline), 5th grade, and 6th grade.

Measures

Screen Time

The latent variable of screen time was comprised of four survey items that included the following prompt: on a regular school day, how many hours do you spend (ST1) on a computer at home or away from school, (ST2) watching television or video movies at home or away from school, (ST3) playing video games that you sit down and play like PlayStation, Xbox, Game Boy, or arcade games, (ST4) playing video games that make you move or breathe hard like Nintendo Wii?

Anhedonia

The latent variable of anhedonia was assessed with four items extracted from the Center for Epidemiologic Studies Depression (CES-D), which is a 20-item self-report scale used to assess symptoms of depression based on symptoms defined by the American Psychiatric Association Diagnostic and Statistical Manuel (DSM-V) for a major depressive episode. The CES-D has a 4-factor structure that includes factors of anhedonia, negative affect, somatization, and interpersonal challenges (Radloff, 1977). Factor analysis of the CES-D has repeatedly shown a 4-factor structure among numerous studies (Shafer, 2006). The four items that load onto the latent variable include: (ANH1) I felt that I was just as good as other people, (ANH2) I felt hopeful about the future, (ANH3) I was happy, and (ANH4) I enjoyed life. Student participants were prompted to reflect on these statements and select one of the following options corresponding to their experience of each item within the past 2 weeks: rarely or none of the time (0–1 day); some or a little of the time (2–3 days); occasionally or a moderate amount of time (4–5 days); and most or all of the time (6–7 days). The composite mean score of these items has demonstrated structural and discriminant validity among non-clinical adolescent and adult populations and has been used to reliably predict substance use outcomes, such as smoking cessation (Leventhal, Ramsey, Brown, LaChance, & Kahler, 2008; Leventhal, Ray, Rhee, & Unger, 2012; Leventhal & Zvolensky, 2015; Shafer, 2006). These four items were reverse-coded so that higher scores indicated greater anhedonia and lack of hedonic capacity.

Substance Use

The latent variable of substance use was assessed with two items adapted from previous studies (Johnston, O’Malley, & Bachman, 2002; Pentz et al., 1989a; Pentz et al., 1989b) that have assessed lifetime substance use among adolescents. These items included (ALC) have you ever tried alcohol (beer, wine, liquor that is not for religious purposes) in your whole life and (CIG) have you ever tried a cigarette in your whole life. Responses for the alcohol item included the following options: (1) no, not even a sip, (2) yes, just a sip, and (3) yes, more than a sip. Responses for the cigarette item included the following options: (1) no, not even a puff, (2) yes, just a puff, (3) yes, more than a puff, and (4) yes, one cigarette or more. Mean composite scores for each item were calculated and loaded onto the latent variable of substance use.

Covariates

Gender was included as a moderator in the model. Race/ethnicity, individual substance use, and reported peer substance use at baseline (Wave A) were included as covariates within the analyses. Peer substance is a strong predictor of individual substance use given emerging effects of socialization in early adolescence ((Bahr, Hoffmann, & Yang, 2005; D’Amico & McCarthy, 2006; Simons-Morton & Chen, 2006). Gender was dichotomously coded as male and female. Participants self-reported their race/ethnicity with the following options: white, black or African American, Hispanic/Latino, Asian, mixed/bi-Racial, or other. Race/ethnicity was dichotomously coded in the analyses as white and non-white for ease of interpretation. Rates of substance use among various different races and ethnicities differ substantially during adolescence and then dissipate in early adulthood, with Hispanics and African Americans reporting greater substance use in early adolescence compared to white adolescents (Chen & Jacobson, 2012; Kuhn, 2015). Despite differences in substance use rates, we opted not to explore this variable as a moderator, since there have been no previous studies that have reported race- and ethnicity-related differences in hours of screen time or levels of anhedonia among early adolescents. Mean composite scores of baseline individual substance use and reported peer substance use (including items: how many of your closest 5 friends have initiated alcohol and cigarette use) were calculated for both items, respectively. Individual substance use at baseline was included in the model as a covariate to control for past substance use (prior to the 4th grade). Similar to previous statistical methodology employed in the parent study (Pentz & Riggs, 2013), preliminary analyses revealed no significant differences between intervention and control groups on our variables of interest, including demographics, screen time, anhedonia, and substance use. Therefore, the intervention of the parent trial was not included as a covariate in the model.

Data Analysis

Descriptive statistics and variables of interest

Means and frequencies of variable of interest were computed. Independent samples t-tests were utilized to test for mean differences in variables of interest between boys and girls. Chi-squared tests were utilized to test for associations between categorical variables.

Structural Equation Modeling (SEM) and FIML Approach

Structural equation modeling (SEM) including full information maximum likelihood (FIML) estimation was conducted using the program EQS 6.4 for Windows (Byrne, 1994). FIML estimation was used to handle missing data, as this method provides less biased estimates compared to other procedures, such as multiple imputation and listwise deletion, for SEM analyses (Schafer, 1977; Olinsky, Chen, & Harlow, 2003). SEM enables mediation and moderation analyses involving latent variables with a reduction in bias while statistically controlling for measurement error (Gunzler et al., 2013; Peyrot, 1996). The incorporation of latent variables allows for ease of interpretation of underlying constructs, as variable items that group together are considered to reflect a unifying factor that may not have been directly measured. Unlike simple regression analysis, SEM also produces model fit statistics, which provide insight onto whether the proposed model fits the data well and whether a theorized causal model can be substantiated (Gunzler et al., 2013; Peyrot, 1996).

Associations between individual items and their respective latent variable were assessed utilizing confirmatory factor analyses (CFA). Each item was expected to have a loading equal or greater than 0.4 and be significantly associated with its corresponding latent variable. Overall model fit was then assessed by examining the following fit statistics: goodness-of-fit χ2 test statistic, comparative fit index (CFI), and root mean squared error of approximation (RMSEA). A CFI of over .9 is considered a good fit (Browne, 1993). A RMSEA value of less than .05 indicates a close fit. A RMSEA value between .05 to .08 and a SRMR value of less than .08 are considered adequate for model fit (Kline, 2005). Distributions of the variables were assessed for normality using skewness and kurtosis statistics. The Lagrange Modifier (LM) test was utilized to assess parameters and their impact on model fit. Only error covariances that were theoretically sound were included to maintain the validity of the final model.

Regression pathways were investigated to assess the structural mediation path between anhedonia, screen time, and substance use. To test for moderation effects of gender, multiple-group SEM approach was utilized that incorporated a two-step process to test for model invariance, including (1) testing for invariance of factor loadings and (2) testing for invariance of regression weights using model chi-square tests (Peyrot, 1996). Pathways between items and their respective factors were constrained to be equal among the two groups indicating that factor loadings could be treated the same. After obtaining acceptable model fit, corresponding regression weights between female and male groups were also constrained to be equal. Chi-square tests were used to determine the significance level of the change in model fit. Retainment of model fit and a non-significant difference in chi-square after constraining the regression weights would indicate that equality constraints are adequate, and no moderation effect has manifested.

Results

Sample Demographics and Missing Data

Demographics and bivariate analyses of variables of interest between boys and girls are provided in Table 1; results revealed significant differences between boys and girls in daily hours of screen time, alcohol use, and cigarette use. Boys spent more time in front of a screen compared to girls (M=2.954 (SD=.075) versus M=2.171 (SD=.046), p<.001) and more boys had initiated alcohol (44.48% versus 30.15%, p<.001) and cigarette use (9.03% versus 3.94%, p=.008).

Table 1.

Demographics and Variables of Interest by Gender (N=709)

Variable Name Boys (N=354)
Mean (SD) or N (%)		 Girls (N=355)
Mean (SD) or N (%)		 p-value
Age
4th grade 9.29 (.49) 9.24 (.45) NS
5th grade 10.74 (.54) 10.70 (.51)
6th grade 11.60 (.56) 11.55 (.53)
Race/Ethnicity
White 124 (35.03) 109 (30.70)
African American 11 (3.11) 4 (1.13)
Hispanic/Latino 91 (25.71) 91 (25.63) NS
Asian American 25 (7.06) 31 (8.73)
Multi-Racial 103 (29.10) 120 (33.80)
Screen Timea (hours daily) 2.954 (.075) 2.171 (.046) p<.001
Anhedoniaa (1(min) – 4(max)) 1.944 (.042) 1.896 (.738) NS
Substance Use Initiation Alcoholb
1=No, not even a sip 196 (55.52) 248 (69.86)
2=Yes, just a sip 120 (33.99) 92 (25.92) p<.001
3=Yes, more than a sip 37 (10.49) 15 (4.23)
Cigarettesb
1=No, not even a puff 322 (90.96) 341 (96.06)
2=Yes, just a puff 18 (5.08) 12 (3.38) p=.008
3=Yes, more than a puff 8 (2.26) 0 (.00)
4=Yes, one cig or more 6 (1.69) 2 (.56)
Open in a new tab
a

Independent samples t-test

b

chi-square tests

NS = Not significant

CFA and SEM Results

The results of CFA indicated that all factor loadings were significant, and most items loaded high on their respective factors. Anhedonia included the following factor loadings for each item [AH1 (.44), AH2 (.54), AH3 (.68), AH4 (.68)]. Screen time included the following factor loadings for each item [ST1(.57), ST2 (.46), ST3 (.85), ST4 (.67)] and substance use included the following factor loadings [CIG (.40), ALC (1.00)]. Cronbach’s alpha scores were .764 for screen time (ST1-ST4) and .750 for anhedonia (AH1-AH4). All variables showed normal distributions, and error covariances consistent with theory within the same factor were included following LM tests. Covariates, including gender, race/ethnicity, individual substance use, and reported peer substance use during the 4th grade (baseline), were included in the model. Longitudinal model parameters, including standardized and unstandardized regression weights and standardized errors, are provided in Table 2.

Table 2.

Longitudinal Model Parameters (Standardized and Unstandardized Regression Weights and Standard Errors)

Observed Variable Latent Construct β B SE
ANH1 .426 1.00
ANH2 Anhedonia .529 1.186 .281
ANH3 .814 1.486 .146
ANH4 .767 1.441 .140
ST1 .590 1.00
ST2 Screen Time .483 0.646 .060
ST3 .896 1.573 .134
ST4 .688 1.123 .110
CIG Substance Use Initiation .504 1.00
ALC .778 2.289 .281
Open in a new tab

All listed paths are significant at p<.05

β = standardized coefficients, B= unstandardized coefficients, SE= standard errors

SEM analysis revealed that screen time was significantly associated with anhedonia (β=.161, p<.05) and anhedonia was significantly associated with substance use (β= .148, p<.05). After including anhedonia in the model, screen time and substance use did not reach significance, indicating a complete mediation effect of anhedonia on the association between screen time and substance use. Gender and individual substance use were the only significant covariates in the final mediation model [(β=.101, p<.05) and (β=.480, p<.05), respectively]. Figure 1 depicts the SEM model.

Figure 1.

Figure 1.

Open in a new tab

SEM Model with Standardized Beta Estimates (N=709)

Note: ** Significant at 0.05 level Covariates included in all regression pathways: Gender, Race/Ethnicity, Peer Substance Use, and Individual Substance Use Initiation in 4th grade

Testing for Moderation Effects

Testing for Invariance of Factor Loadings

To initiate multiple group SEM, a base model (M0) combining boys and girls yielded the following fit statistics: χ2 = 202.372 (df=102), p<.001, CFI=.951, RMSEA=.05, SRMR=.044. Applying equality constraints on the factor loadings (model (M1)) resulted in the following fit indices: χ2 = 237.292 (df=109), p<.001, CFI=.936, RMSEA=.055, SRMR=.052. There was a significant difference in chi squares between the two models (Δχ2=34.92, p<.001); therefore, inappropriate constraints based on LM tests were removed. A partially constrained model was retained (Model 1a (M1a)). The fit statistics for M1a were: χ2 = 209.323 (df=107), p<.001, CFI=.949, RMSEA=.050, SRMR=.046.

Testing for Invariance of Regression Weights

To test for moderation effects of gender, all regression weights were constrained to be equal (Model 2), revealing the following fit indices: χ2 = 216.219 (df=110), p<.001, CFI=.947, RMSEA=.050, SRMR=.049. A chi-square difference test revealed no significant difference in chi-squares between model 2 and model 1a (Δχ2=6.896, p=.0725); therefore, the fully constrained model 2 was retained as the final model. This non-significant difference in model fit indicated there was no moderation effect of gender within the model. Model development and fit indices are shown in Table 3.

Table 3.

Model Development and Fit Statistics for Multiple Group SEM Approach

χ2 df p-value CFI RMSEA
BOYS
MB1 130.074 53 <.001 .932 .061
MB2 100.018 51 <.001 .958 .049
GIRLS
MG1 175.976 53 <.001 .856 .080
MG2 102.354 51 <.001 .941 .052
COMBINED χ2 df p-value CFI RMSEA SRMR
M0 202.372 102 <.001 .951 .050 .044
M1 237.292 109 <.001 .936 .055 .052
M1a 209.323 107 <.001 .949 .050 .046
M2 216.219 110 <.001 .947 .050 .049
Open in a new tab

Note: MB1 and MG1= basic theoretical models for boys and girls. MB2 and MG2 represent modified models with correlated errors following LM test recommendations. M0 represents the base model of combined groups; M1 represents the base model with all factor loadings constrained; M1a represents partially constrained factor loadings; M2 represents partially constrained factor loadings and fully constrained regression weights

Discussion

Our study addresses a critical gap in existing literature by revealing the important mediation effects of anhedonia on the association between screen time and substance use among early adolescents. To our knowledge, no studies to date have assessed the association between screen time and anhedonia and its relation to substance use behaviors among early adolescents. Although more longitudinal studies that can explain the effect of screen time on anhedonia are needed, we offer some speculative explanations for the significant mediation effects of anhedonia on the pathway between screen time and substance use.

Activities that produce rapid and easy rewards at minimal effort, such as watching television, using a computer, and playing video games, may be associated with anhedonia, categorized as the reduced initiative to pursue activities once found to be enjoyable (Treadway, Buckholtz, Schwartzman, Lambert, & Zald, 2009) (Guillot et al., 2016). Similar to desensitization after repeated exposure to screen violence (Anderson et al., 2017), adolescents may be experiencing similar desensitization to screen rewards. Over time, adolescents may exhibit a blunted response to hedonic effects following repeated exposure to pleasing and rewarding stimuli offered through screen devices. Increased screen time may ultimately lead to more anhedonia among youth and thus greater risk for substance use through the need to compensate for the reduced experience of rewards (Sussman & Leventhal, 2014; Leventhal et al., 2008; Leventhal et al., 2012; Leventhal & Zvolensky, 2015; Shafer, 2006). Additionally, researchers have reported that adolescents who report higher levels of anhedonia are more likely to engage in compulsive, hedonically-driven behaviors that are of higher intensity and risk, such as initiating and escalating cigarette smoking (Sussman & Leventhal, 2014). More time spent in front of screens may exacerbate anhedonic symptoms over time, ultimately leading to engagement in compulsive behaviors to attain rewards and compensate for inherent deficits in the experience of pleasure. Thus, the lack of hedonic response coupled with the tendency to engage in compulsive behaviors to cope with associated deficits may propel anhedonic adolescents to experiment with alcohol and cigarettes.

Although there were no moderation effects of gender in this study, gender was still a significant covariate in our final mediation model. While preliminary analyses revealed that there were no significant gender differences in anhedonia, boys spent significantly more time in front of screens compared to girls, specifically watching television and playing video games. A possible reason for this finding is that anhedonia may be sensitive to specific thresholds of screen time. For instance, it is possible that screen time above a certain threshold results in similar levels of anhedonia among boys and girls. Another reason for this observed difference in hours of screen time between boys and girls may be attributed to increased video game playing among boys. A recent study by Van Rooij and colleagues (2014) found that boys were much more likely to be involved in addictive video gaming behavior and that this behavior was correlated with higher depressive symptomology scores and substance use behavior (tobacco, alcohol, and cannabis)(Van Rooij et al., 2014). The previously mentioned longitudinal study assessing the effects of risk-glorifying video games and behavioral deviance found that boys were significantly more likely to play these games and demonstrated greater risk for alcohol and cigarette use (Hull et al., 2014). Therefore, boys might be increasing their susceptibility to the onset of substance use through exposure to video game content and more hours spent in front of a screen. Additionally, during moments of internal conflict, boys rely more on passivity, avoidance, and suppression, while girls tend to rely on social support and engage in dysfunctional rumination (Zimmermann & Iwanski, 2014). In general boys have the propensity to suppress their negative affective states and thus may spend more time watching television as an attempt to manage their emotions versus alternative strategies, such as seeking support from their peers (Elhai et al., 2018). Although our study did not find moderation effects of gender, it is important to highlight gender differences in screen time exposure reported in the findings, particularly when designing gender-specific substance use prevention programs.

There are several limitations associated with this study. We only used 4 items from the CES-D to comprise the latent variable of anhedonia and did not use anhedonia-specific, self-report scales (Olivares, Berrios, & Bousoño, 2013). We also could not adjust our analyses for other symptoms of depression. Although anhedonia is not highly associated with other depressive symptoms (Leventhal & Zvolensky, 2015), statistical adjustment to account for possible shared variance is necessary to delineate the true mediation effects of anhedonia on the association between screen time and substance use. Studies that assess these constructs in the future would benefit from controlling for other symptoms of depression to attain more robust results. Although we attempted to control for baseline variables of interest, we could not control for baseline anhedonia, as survey times for anhedonia were introduced to the study during the 5th grade. Therefore, we cannot fully disentangle the effects of baseline anhedonia levels when assessing these linear associations. Although anhedonia levels were not expected to change drastically from the 4th to 5th grades, we continue to speculate that screen time may lead to reduced hedonic capacity among adolescents. Only two items were used to construct the latent variable of lifetime substance use (alcohol and cigarette use). Although substance use is a multifactorial construct, use of other substances (inhalants, opioids, heroin) among early adolescents within the United States is rare and reports of alcohol and cigarette use are most prevalent within this age group (Lipari, Ahrnsbrak, Pemberton, & Porter, 2017). Although we focused on substances that were more applicable to our population sample, the incorporation of more factors would have ideally strengthened our latent variable; therefore, future studies would benefit from the inclusion of more items (particularly vaping and cannabis). We also did not assess for specific screen-related content and simply defined screen time as the number of hours spent in front of a screen outside of school. It is still uncertain whether it is the amount of screen time, specific content exposure, or a combination of both that drive these interrelationships.

This study utilized survey data from 2009–2011, representing a time when technology was not as advanced and most elementary school students did not have access to smart phones, particularly disadvantaged students coming from underprivileged backgrounds. Important conceptual insights from these findings can be made on the early harmful effects of screen time before these technological devices were readily available to youth, particularly since the use of screen devices has grown substantially over the last few years. Future studies would benefit from assessing the content of screen exposure when investigating these latent variables. Investigators should determine what constitutes pathological behaviors regarding screen time, including assessment of internet addiction disorders or analytic comparisons between adolescents who spend a greater amount of time in front of the screen compared to adolescents who spend less time. While we chose to focus on substance use for this study, assessment of how many hours might lead to pathological addiction regarding the use of technology and screen time may provide insight on how to prevent these disorders among youth.

With recent technological development of innovative screen devices, it is critical to evaluate the effects of increased screen time on adolescent health-related outcomes, not solely focusing on physical health determinants. This longitudinal study can provide insight on the important role of anhedonia on the effects of screen time on substance use. Future studies can focus on protective factors that can combat the adverse effects of screen time on substance use among adolescents. Additionally, interventions designed to prevent substance use may benefit from targeting anhedonic youth, who may represent a particularly vulnerable population.

Acknowledgments

Funding: This manuscript was written with the support of a National Cancer Institute T32 grant to the first author [T32CA009492-33](PI: M.P.). Data were provided from the National Institutes of Health Grant [R01HD052107-0182] (PI: M.P.)

Footnotes

Conflicts of Interest: The authors declare no conflicts of interest with this study.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Age limits and adolescents. (2003). Paediatr Child Health, 8, 577–578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andersen SL, & Teicher MH (2008). Stress, sensitive periods and maturational events in adolescent depression. Trends in Neurosciences, 31, 183–191. [DOI] [PubMed] [Google Scholar]
  3. Anderson CA, Bushman BJ, Bartholow BD, Cantor J, Christakis D, Coyne SM,… Ybarra M (2017). Screen Violence and Youth Behavior. Pediatrics, 140(Suppl 2), S142–S147. [DOI] [PubMed] [Google Scholar]
  4. Anderson CA, Economos CD, Must A (2008). Active play and screen time in US children aged 4 to 11 years in relation to sociodemographic and weight status characteristics: a nationally representative cross-sectional analysis. BMC Public Health, 8, 366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bahr SJ, Hoffmann JP, & Yang X (2005). Parental and peer influences on the risk of adolescent drug use. Journal of Primary Prevention, 26, 529–551. [DOI] [PubMed] [Google Scholar]
  6. Bechara A (2005). Decision making, impulse control and loss of willpower to resist drugs: a neurocognitive perspective. Nat Neurosci, 8, 1458–1463. [DOI] [PubMed] [Google Scholar]
  7. Bennett DS, Ambrosini PJ, Kudes D, Metz C, & Rabinovich H (2005). Gender differences in adolescent depression: do symptoms differ for boys and girls? Journal of Affective Disorders, 89, 35–44. [DOI] [PubMed] [Google Scholar]
  8. Boers E, Afzali MH, Newton N, & Conrod P (2019). Association of Screen Time and Depression in Adolescence. JAMA Pediatrics, 173, 853–859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Browne M, & Cudeck R (1993). Alternative ways of assessing model fit . In Bollen K, Long J (Ed.), Testing structural equation models (pp. 136–162). Newbury Park: Sage Publications. [Google Scholar]
  10. Byrne BM (1994). Structural equation modeling with EQS and EQS/Windows: basic concepts, applications, and programming. California: Sage Publications. [Google Scholar]
  11. Cano MA, de Dios MA, Correa-Fernandez V, Childress S, Abrams JL, & Roncancio AM (2017). Depressive symptom domains and alcohol use severity among Hispanic emerging adults: Examining moderating effects of gender. Addictive Behaviors, 72, 72–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Carleton RN, Thibodeau MA, Teale MJ, Welch PG, Abrams MP, Robinson T, & Asmundson GJ (2013). The center for epidemiologic studies depression scale: a review with a theoretical and empirical examination of item content and factor structure. PLoS One, 8, e58067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Carson V, Pickett W, & Janssen I (2011). Screen time and risk behaviors in 10- to 16-year-old Canadian youth. Preventive Medicine, 52, 99–103. [DOI] [PubMed] [Google Scholar]
  14. Casey BJ, Jones RM, & Hare TA (2008). The adolescent brain. Annals of the New York Academy of Sciences, 1124, 111–126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Chen P, & Jacobson KC (2012). Developmental trajectories of substance use from early adolescence to young adulthood: gender and racial/ethnic differences. Journal of Adolescent Health, 50, 154–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Chuang CW, Chan C, & Leventhal AM (2016). Adolescent Emotional Pathology and Lifetime History of Alcohol or Drug Use With and Without Comorbid Tobacco Use. Journal of Dual Diagnoses, 12, 27–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Cuthbert BN, & Insel TR (2013). Toward the future of psychiatric diagnosis: the seven pillars of RDoC. BMC Medicine, 11, 126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. D’Amico EJ, & McCarthy DM (2006). Escalation and initiation of younger adolescents’ substance use: the impact of perceived peer use. Journal of Adolescent Health, 39, 481–487. [DOI] [PubMed] [Google Scholar]
  19. Elhai JD, Tiamiyu MF, Weeks JW, Levine JC, Picard KJ, & Hall BJ (2018). Depression and emotion regulation predict objective smartphone use measured over one week. Personality and Individual Differences, 133, 21–28. [Google Scholar]
  20. Foltz JL, Cook SR, Szilagyi PG, Auinger P, Stewart PA, Bucher S, & Baldwin CD, (2011). US adolescent nutrition, exercise, and screen time baseline levels prior to national recommendations. Clinical Pediatrics, 50, 424–33. [DOI] [PubMed] [Google Scholar]
  21. Garfield JB, Lubman DI, & Yucel M (2014). Anhedonia in substance use disorders: a systematic review of its nature, course and clinical correlates. Australian and New Zealand Journal of Psychiatry, 48, 36–51. [DOI] [PubMed] [Google Scholar]
  22. Guillot CR, Bello MS, Tsai JY, Huh J, Leventhal AM, & Sussman S (2016). Longitudinal Associations between Anhedonia and Internet-Related Addictive Behaviors in Emerging Adults. Computational Human Behavior, 62, 475–479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gunzler D, Chen T, Wu P, & Zhang H (2013). Introduction to mediation analysis with structural equation modeling. Shanghai Arch Psychiatry, 25(6), 390–394. doi: 10.3969/j.issn.1002-0829.2013.06.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ha JH, Kim SY, Bae SC, Bae S, Kim H, Sim M,… Cho SC (2007). Depression and Internet addiction in adolescents. Psychopathology, 40(6), 424–430. [DOI] [PubMed] [Google Scholar]
  25. Hands BP, Chivers PT, Parker HE, Beilin L, Kendall G, & Larkin D (2011). The associations between physical activity, screen time and weight from 6 to 14 yrs: the Raine Study. Journal of Science and Medicine in Sport, 14, 397–403. [DOI] [PubMed] [Google Scholar]
  26. Hanewinkel R, Sargent JD, Poelen EAP, Scholte R, Florek E, Sweeting H,. . . Morgenstern M (2009). Alcohol Consumption in Movies and Adolescent Binge Drinking in 6 European Countries. Pediatrics, 129, 709–719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hoare E, Milton K, Foster C, & Allender S (2016). The associations between sedentary behaviour and mental health among adolescents: a systematic review. International Journal of Behaviroal Nutrition and Physical Activity, 13, 108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hull JG, Brunelle TJ, Prescott AT, & Sargent JD (2014). A longitudinal study of risk-glorifying video games and behavioral deviance. Journal of Personality and Social Psychology, 107, 300–325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Husain M, & Roiser JP (2018). Neuroscience of apathy and anhedonia: a transdiagnostic approach. Nature Reviews Neurosciences, 19, 470–484. [DOI] [PubMed] [Google Scholar]
  30. Johnston LD, O’Malley PM & Bachman JG (2002) Monitoring the Future National Survey Results on Drug Use, 1975–2001, Vol. 1 Secondary School Students. NIH Publications no. 02–5106. Bethesda, MD: US Department of Health and Human Services, National Institute on Drug Abuse [Google Scholar]
  31. Khesht-Masjedi MF, Shokrgozar S, Abdollahi E, Golshahi M, & Sharif-Ghaziani Z (2017). Comparing depressive symptoms in teenage boys and girls. Journal of Family Medicine and Primary Care, 6, 775–779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kline RB (2005). Principles and practice of structural equation modeling. New York: Guilford Press. [Google Scholar]
  33. Ko CH, Yen JY, Yen CF, Chen CS, Weng CC, & Chen CC (2008). The association between Internet addiction and problematic alcohol use in adolescents: the problem behavior model. Cyberpsychology and Behavior, 11, 571–576. [DOI] [PubMed] [Google Scholar]
  34. Kuhn C (2015). Emergence of sex differences in the development of substance use and abuse during adolescence. Pharmacology & Therapeutics, 153, 55–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Leventhal AM, Ramsey SE, Brown RA, LaChance HR, & Kahler CW (2008). Dimensions of depressive symptoms and smoking cessation. Nicotine & Tobacco Research, 10, 507–517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Leventhal AM, Ray LA, Rhee SH, & Unger JB (2012). Genetic and environmental influences on the association between depressive symptom dimensions and smoking initiation among Chinese adolescent twins. Nicotine & Tobacco Research, 14, 559–568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Leventhal AM, Strong DR, Sussman S, Kirkpatrick MG, Unger JB, Barrington-Trimis JL, & Audrain-McGovern J (2016). Psychiatric comorbidity in adolescent electronic and conventional cigarette use. Journral of Psychiatric Research, 73, 71–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Leventhal AM, Unger JB, Audrain-McGovern J, Sussman S, Volk HE, & Strong DR (2015). Measuring Anhedonia in Adolescents: A Psychometric Analysis. Journal of Personality Assessment, 97, 506–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Leventhal AM, & Zvolensky MJ (2015). Anxiety, depression, and cigarette smoking: a transdiagnostic vulnerability framework to understanding emotion-smoking comorbidity. Psychological Bulletin Journal, 141, 176–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lipari RN, Ahrnsbrak RD, Pemberton MR, Porter JD (2017). Risk and Protective Factors and Estimates of Substance Use Initiation: Results from the 2016 National Survey on Drug Use and Health. CBHSQ Data Review. Rockville, MD: Substance Abuse and Mental Health Services Administration (US). [PubMed] [Google Scholar]
  41. Maras D, Flament MF, Murray M, Buchholz A, Henderson KA, Obeid N, & Goldfield GS (2015). Screen time is associated with depression and anxiety in Canadian youth. Preventive Medicine, 73, 133–138. [DOI] [PubMed] [Google Scholar]
  42. McMakin DL, Olino TM, Porta G, Dietz LJ, Emslie G, Clarke G,… Brent DA (2012). Anhedonia predicts poorer recovery among youth with selective serotonin reuptake inhibitor treatment-resistant depression. Journal of the American Academy of Child & Adolescent Psychiatry, 51, 404–411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Nelson SE, van Ryzin MJ, & Dishion TJ (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. [DOI] [PubMed] [Google Scholar]
  44. Ohannessian CM (2009). Media Use and Adolescent Psychological Adjustment: An Examination of Gender Differences. Journal of Child & Family Studies, 18, 582–593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Olinsky A, Chen S, & Harlow L (2003). The comparative efficacy of imputation methods for missing data in structural equation modeling. European Journal of Operational Research, 151, 53–79. [Google Scholar]
  46. Olivares JM, Berrios GE, Bousoño M (2013). The self-assessment anhedonia scale (SAAS). Neurology, Psychiatry, & Brain Research,12, 121–134. [Google Scholar]
  47. Pang RD, Farrahi L, Glazier S, Sussman S, & Leventhal AM (2014). Depressive symptoms, negative urgency and substance use initiation in adolescents. Drug and Alcohol Dependence, 144, 225–230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Park S, Hong KE, Park EJ, Ha KS, & Yoo HJ (2013). The association between problematic internet use and depression, suicidal ideation and bipolar disorder symptoms in Korean adolescents. Australian and New Zealand Journal of Psychiatry, 47, 153–159. [DOI] [PubMed] [Google Scholar]
  49. Pentz MA, Dwyer JH, MacKinnon DP, Flay BR, Hansen WB, Wang EY, & Johnson CA (1989a). A multicommunity trial for primary prevention of adolescent drug abuse. Effects on drug use prevalence. JAMA, 261, 3259–3266. [PubMed] [Google Scholar]
  50. Pentz MA, Johnson A, Dwyer JH, MacKinnon DM, Hansen WB, & Flay BR (1989b). A comprehensive community approach to adolescent drug abuse prevention: effects on cardiovascular disease risk behaviors. Annals of Medicine, 21, 219–222. [DOI] [PubMed] [Google Scholar]
  51. Pentz MA, & Riggs NR (2013). Longitudinal relationships of executive cognitive function and parent influence to child substance use and physical activity. Prevention Science, 14, 229–237. [DOI] [PubMed] [Google Scholar]
  52. Peyrot M (1996). Causal analysis: theory and application. Journal of Pediatric Psychology, 21, 3–24. [DOI] [PubMed] [Google Scholar]
  53. Primack BA, Kraemer KL, Fine MJ, & Dalton MA (2009). Media exposure and marijuana and alcohol use among adolescents. Substance Use & Misuse, 44, 722–739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Radloff LS (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement. Applied Psychological Measurement, 1, 385–401. [Google Scholar]
  55. Rucker J, Akre C, Berchtold A, & Suris JC (2015). Problematic Internet use is associated with substance use in young adolescents. Acta Paediatrica, 104, 504–507. [DOI] [PubMed] [Google Scholar]
  56. Sakuma KL, Riggs NR, & Pentz MA (2012). Translating evidence based violence and drug use prevention to obesity prevention: development and construction of the pathways program. Health Education Ressearch, 27, 343–358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Saluja G, Iachan R, Scheidt PC, Overpeck MD, Sun W, & Giedd JN (2004). Prevalence of and risk factors for depressive symptoms among young adolescents. Archives of Pediatric & Adolescent Medicine, 158, 760–765. [DOI] [PubMed] [Google Scholar]
  58. Saunders TJ, & Vallance JK (2017). Screen Time and Health Indicators Among Children and Youth: Current Evidence, Limitations and Future Directions. Appiedl Health Economics and Health Policy, 15, 323–331. [DOI] [PubMed] [Google Scholar]
  59. Schafer JL (1977). Analysis of incomplete multivariate data. London: Chapman & Hall. [Google Scholar]
  60. Shafer AB (2006). Meta-analysis of the factor structures of four depression questionnaires: Beck, CES-D, Hamilton, and Zung. Journal of Clinical Psychology, 62, 123–146. [DOI] [PubMed] [Google Scholar]
  61. Shin HS, Valente TW, Riggs NR, Huh J, Spruijt-Metz D, Chou CP, & Ann Pentz M (2014). The interaction of social networks and child obesity prevention program effects: the pathways trial. Obesity (Silver Spring), 22, 1520–1526. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Simons-Morton B, & Chen RS (2006). Over time relationships between early adolescent and peer substance use. Addictive Behaviors, 31, 1211–1223. [DOI] [PubMed] [Google Scholar]
  63. Sisson SB, Church TS, Martin CK, Tudor-Locke C, Smith SR, Bouchard C,… Katzmarzyk PT (2009). Profiles of sedentary behavior in children and adolescents: the US National Health and Nutrition Examination Survey, 2001–2006. International Journal of Pediatric Obesity, 4, 353–359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Smith KL, Hummer TA, & Hulvershorn LA (2015). Pathological Video Gaming and Its Relationship to Substance Use Disorders. Current Addiction Reports, 2, 302–309. [Google Scholar]
  65. Stueve A, & O’Donnell LN (2005). Early alcohol initiation and subsequent sexual and alcohol risk behaviors among urban youths. American Jounral of Public Health, 95, 887–893. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Sussman S, & Leventhal A (2014). Substance misuse prevention: addressing anhedonia. New Directions for Youth Development, 10, 45–56, [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Treadway MT, Buckholtz JW, Schwartzman AN, Lambert WE, & Zald DH (2009). Worth the ‘EEfRT’? The effort expenditure for rewards task as an objective measure of motivation and anhedonia. PLoS One, 4, e6598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Twenge JM, & Campbell WK (2018). Associations between screen time and lower psychological well-being among children and adolescents: Evidence from a population-based study. Preventive Medicine Reports, 12, 271–283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Twenge JM, Joiner TE, Rogers ML, & Martin GN (2018). Increases in Depressive Symptoms, Suicide-Related Outcomes, and Suicide Rates Among U.S. Adolescents After 2010 and Links to Increased New Media Screen Time. Clinical Psychological Science, 6, 3–17. [Google Scholar]
  70. Van Rooij AJ, Kuss DJ, Griffiths MD, Shorter GW, Schoenmakers MT, & Van De MD (2014). The (co-)occurrence of problematic video gaming, substance use, and psychosocial problems in adolescents. Journal of Behavioral Addictions, 3, 157–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Waller MW, Hallfors DD, Halpern CT, Iritani BJ, Ford CA, & Guo G (2006). Gender differences in associations between depressive symptoms and patterns of substance use and risky sexual behavior among a nationally representative sample of U.S. adolescents. Archives of Women’s Mental Health, 9, 139–150. [DOI] [PubMed] [Google Scholar]
  72. Welty LJ, Hershfield JA, Abram KM, Han H, Byck GR, & Teplin LA (2017). Trajectories of Substance Use Disorder in Youth After Detention: A 12-Year Longitudinal Study. Jounral of American Academy of Child & Adolescent Psychiatry, 56, 140–148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Zimmermann P, & Iwanski A (2014). Emotion regulation from early adolescence to emerging adulthood and middle adulthood Age differences, gender differences, and emotion-specific developmental variations. International Journal of Behavioral Development, 38, 182–194. [Google Scholar]

RESOURCES