Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Subst Use Misuse. Author manuscript; available in PMC 2017 April 15.
Published in final edited form as:
PMCID: PMC4861039

Separating the Association between Inhibitory Control and Substance Use Prevalence versus Quantity during Adolescence: A hurdle mixed-effects model approach



Inhibitory control is a critical component to the self-regulation of affect and behavior. Research consistently demonstrates negative associations between inhibitory control and several problem behaviors including substance misuse during early adolescence. However, analytic approaches previously used have often applied ordinary least squares (OLS) regression to non-normal count data with an excessive number of zeros (i.e., never users), violating several model assumptions. Further, OLS regression fails to model effects of the independent variable, separately, for both prevalence and quantity of use.


The study objective was to simultaneously model associations between inhibitory control and both past 30- day prevalence and amount of cigarette and marijuana use. It was hypothesized that when doing so, inhibitory control would be significantly associated with prevalence, but not quantity of use.


Hurdle Mixed-effects Models (HMM) were used for hypothesis testing on data collected from 3,383, 9th grade adolescents (Mage = 14.08 years).


Results confirmed hypotheses, demonstrating that although significant bivariate associations between inhibitory control and quantity of cigarette and marijuana use existed, HMM analyses established that the associations were more precisely specific to past 30-day prevalence, and not quantity of use.


Results from a HMM approach contribute to a more comprehensive and nuanced understanding of which characteristics of cigarette and marijuana use are associated with inhibitory control during early adolescence.

Keywords: Inhibitory control, cigarette use, marijuana use, prevalence, hurdle mixed-effects models


Inhibitory control, defined as the capacity to inhibit a prepotent or automatic thought or action in favor of a more desirable, or healthy response (Davidson, Amso, Anderson, & Diamond, 2006), is central to many models of the self-regulation of problem behavior (de Ridder, Lensvelt-Mulders, Finkenauer, Stok, & Baumeister, 2012). It is considered to be a central component of executive function (EF), the umbrella term for variety of neurocognitive processes necessary for effective problem-solving and goal directed health behaviors (Zelazo, Carlson, Kesek, 2008). Inhibitory control is associated with the developmental integration of the brain’s prefrontal cortex (PFC) and other reward (e.g., mesolimbic dopaminergic system) and affect (e.g., amygdala) centers of the brain (Munakata, Herd, Chatham, Depue, Banich, & O’Reilly, 2011). An important characteristic of this developmental integration is that the areas of the brain responsible for the generation of affective impulses and reward sensitivity achieve structural and functional maturity much earlier in life than does the PFC, which is responsible for secondary processing and regulation of those impulses.

The implication of this protracted time-course of PFC development is that late childhood and adolescence are characterized as periods during which behavior is disproportionately influenced by strong emotional impulses and drives for novel experiences (Chambers, Taylor, & Potenza, 2003). In addition, adolescence is a period of growing social pressure and drives for independence, which creates increasing opportunities to engage in risky behaviors, including substance use (Reyna & Farley 2006). Thus, adolescence is a stage in which youth engage in new levels of emotional stimulation and risk-taking while still having a less than fully developed inhibitory control skills necessary for regulating those impulses (Steinberg, Dahl, Keating, Kupfer, Masten, & Pine, 2006).

Notably, substance use increases during adolescence (Riggs, Chou, Li, & Pentz, 2007) with inhibitory control being one factor consistently shown in cross-sectional and longitudinal studies to be associated with substance misuse (Iacono, Carlson, Taylor, Elkins, & McGue, 1999; Ivanov, Schultz, London, & Newcorn, 2008). For example, Tarter and colleagues demonstrated that the converse of inhibitory control, “neurobehavioral disinhibition,” as measured by interview, questionnaire, and behavioral EF tasks at age 16 years predicted substance use disorder at age 19 (Tarter, Kirisci, Mezzich, Cornelius, Pajer, Vanyukov et al., 2003). Similarly, Nigg and colleagues (2006) demonstrated that poor “response inhibition” predicted illicit drug use in adolescents at risk for substance use disorders. Imaging studies appear to confirm behavioral research illustrating that neural indicators of behavioral disinhibition (i.e., P300 amplitude and decreased regional blood flow to the PFC) have been linked to greater substance use during adolescence (Dom, Sabbe, Hulstijn, & van den Brink, 2005; Moeller, Barratt, Fischer, Dougherty, Reilly, Mathias et al., 2004; Norman, Pulido, Squeglia, Spadoni, Paulus, & Tapert, 2011). Altogether, this research indicates that inhibitory control is associated with the misuse of an array of substances during adolescence (Iacono et al., 1999; Riggs, 2015).

Less clear is the aspect of adolescent substance use (prevalence vs. amount of use) that inhibitory control relates to. Little is known with respect to whether proficient inhibitory control protects youth from initiating substance use, whether it facilitates regulation of the amount of use once initiated, or both, which has implications for prevention tactics aimed at offsetting risk of substance use due to inhibitory control deficits. Initially, inhibitory control may facilitate adolescents’ capacity to refuse substance use offers within a context of peer reinforcement for use (Trucco, Colder, & Wieczorek, 2011). Here, one process by which inhibitory control may be related to substance misuse is that adolescents who are more proficient at inhibiting strong impulses may never initiate use to begin with. Inhibitory control may also enhance the capacity to regulate the amount of substance use intake once initiated, potentially preventing patterns of problematic overuse. That is, proficient inhibitory control skills may aid the prevention of sustained, disinhibited overuse.

One limitation of the previous research in this area is that studies often fail to simultaneously model the prevalence and amount of substance use. Failing to do so renders it difficult to rule out the competing hypothesis that associations between inhibitory control and problematic amounts of substance use are due to maintained associations between inhibitory control and substance use prevalence. Further articulating associations between inhibitory control and patterns of substance use is of research importance so that prevention scientists can more precisely translate and time substance use preventive interventions focusing on promoting conscious strategies for inhibitory control.


The purpose of this study was to simultaneously test associations between inhibitory control and cigarette and marijuana use prevalence and quantity. It was hypothesized that, when simultaneously modeling associations between inhibitory control and 30-day prevalence (any use on at least one day of the past 30 vs. zero days) and amount of use (quantity used on days used in the past 30), inhibitory control would be significantly associated with prevalence, but not amount of cigarette and marijuana use. Significant negative associations between inhibitory control and past 30-day use prevalence were expected due to inhibitory control’s capacity to facilitate inhibition of initial use. However, associations between inhibitory control and quantity of use were not expected for two reasons. First, one’s preexisting capacity to inhibit behavior, whether that be inhibitory proficiency or deficit, may become disrupted once the acute psychoactive effects of substances affect prefrontal cortical function (Bolla, Brown, Eldreth, Tate, & Cadet, 2002; Goldstein & Volkow, 2002; Gonzalez, 2007; Stein, Pankiewicz, Harsch, Cho, Fuller, Hoffman, et al., 1998). That is, once substance use is initiated, the preexisting capacity to inhibit continued use becomes disrupted, resulting in a lack of association between inhibitory control and quantity of use. Second, it was expected, due to the early developmental stage of this community (i.e., non-clinical) sample, that there would not be sufficient use among the vast majority of those who have used to contribute to chronic dysregulation of trait inhibitory control.


Participants and Procedures

Participants were 9th grade (Mage = 14.08, SD = .42) students enrolled in 10 Los Angeles metropolitan area public high schools invited to participate in a study investigating associations between psychopathology and health behavior. The schools were selected based on their adequate representation of diverse demographic characteristics; the percent of students eligible for free lunch within each school (i.e., student’s parental income ≤ 185% of the national poverty level) on average across the ten schools was 31.1% (SD = 19.7, range: 8.0% - 68.2%). All students were eligible to participate with the exception of those in either special education or English as a second language programs (N = 4,100). Of the 3,874 (94.5%) who provided parental consent, 3,383 (82.5%) provided verbal assent, were enrolled in the study, and were administered the study survey. Forty-seven percent of participants were male. Participants were 47% Latino/Hispanic, 16% Caucasian, 16% Asian-American, 5% African-American, with 16% identifying as Native American, Multi-Racial, or an “Other” race.

Participants completed in-person, paper-and-pencil self-report surveys administered by trained research assistants during two 40-minute class periods. Students were informed that their responses were confidential and would not be shared with their teachers, parents, or school staff. Each school was compensated $2,500 for their general activity fund; students were not individually compensated. Certificates of confidentiality issued by the federal government and the University of Southern California’s Institutional Review Board (IRB) were provided to participants to ensure an understanding that only the university’s researchers would have access to their data. All procedures were approved by the university’s IRB.


Inhibitory control was measured using the five-item Inhibitory Control scale of the Early Adolescent Temperament Questionnaire – Revised (EATQ-R) (Capaldi & Rothbart, 1992; Ellis & Rothbart, 2001). Sample items include: “When someone tells me to stop doing something, it is easy for me to stop”; and “The more I try to stop myself from doing something I shouldn’t, the more likely I am to do it.” Item responses ranged from 1 = almost always untrue to 5 = almost always true. The EATQ-R has established internal consistency (alpha = .69), test-retest reliability (r = .78), and convergent validity with parent reports and other constructs indicative of effortful control (Ellis & Rothbart 2001; Muris & Meesters, 2009). Responses from items phrased such that lower scores represented greater inhibitory control were reversed.

Internal consistency for the EATQ-R was lower in the current study, alpha = .49. This low internal consistency may represent problems with measurement reliability. However, it may also mean that the items are tapping relatively distinct components of a broader construct. To address this issue we used an item-deletion approach which resulted in a 3-item inhibitory control scale with alpha = .62. The 3-item scale more closely approximates the internal consistency of the EATQ-R found in previous studies (Ellis et al., 2009), and is likely lower due to the reduced set of items. Models were run for both the 3-item and 5-item EATQ-R scales to demonstrate consistency of results (see Footnote 1).

Past 30-day cigarette and marijuana use items were based on the Youth Behavior Risk Surveillance Survey (YBRS) and Monitoring the Future Questionnaire questions (MTF), which have been extensively validated with adolescents (Eaton, Kann, Kinchen, Shanklin, Ross, Hawkings et al., 2010; Johnston, O’Malley, Bachman, & Schulenberg, 2010). Quantity of cigarette use was assessed by asking youth, “During the past 30 days, on the days you smoked, how many cigarettes did you smoke per day”? Item responses ranged from 0 = I did not smoke cigarettes during the past 30 days to 8 = more than 20 cigarettes per day. Quantity of marijuana use was assessed by asking youth, “During the past 30 days, on the days you smoked marijuana, about how many marijuana cigarettes, joints, reefers, or the equivalent, did you usually have”? Item responses ranged from 0 = I did not smoke marijuana during the past 30 days to 8 = 11 or more a day.

Data Analyses

All analyses were conducted using R version 3.1.2. One primary challenge to better understanding the process by which inhibitory control is associated with substance use is that much of the research conducted in this area employs ordinary least squares (OLS) regression to model associations between inhibitory control and self-reported substance using count data. This is often inappropriate due to two related issues which are particularly problematic during early adolescence: 1) the non-normality and heteroskedasticity of reported use, and 2) the excessive number of zeros (i.e., those who report never using substances). Zero-altered count models, encompassing zero-inflated Poisson, zero-inflated negative binomial, and hurdle models, can be employed to address these issues, in part, by producing two separate models (Atkins, Baldwin, Zheng, Gallop, & Neighbors, 2013). The first is a “count” model employing a logarithmic transformation, much in the same way log transformations of a dependent variable are employed to ameliorate the effects of non-normal residuals. Resulting coefficients can subsequently be exponentiated to facilitate interpretation in a similar fashion as OLS regression. The second model is a logit model predicting membership in the certain zero, or “never used in the past 30 days” category, and can be interpreted as an odds-ratio following transformation.

These models account for overdispersion, when the conditional variance of an outcome variable exceeds the conditional mean, a violation of a fundamental assumption of Poisson regression. This issue becomes more pronounced with the excessive number of zeros often found in substance use data collected during early adolescence. However, the logistic portions of zero-inflated models are taken into account when calculating the count portions such that the logistic portions only model the excess zeros within the data. Hurdle models, in contrast, completely separate zero scores and assess them within the logistic portions, resulting in a truncated count model (i.e., all individuals who used on one or more occasions). These models are best suited for comparing effects within students who have already initiated substance use to effects between users and non-users, and have been used to longitudinally model substance use outcomes (Neighbors, Atkins, Lewis, Lee, Kaysen, Mittmann et al., 2011).

This study employed Bayesian hurdle mixed-effects models (HMM) to simultaneously test the associations between inhibitory control and both substance use prevalence and amount of use. To account for nesting within schools for cigarette (ICC = .001) and marijuana use (ICC = .02), generalized linear mixed models (GLMMs; McCulloch & Neuhause, 2003) were estimated using Bayesian Markov chain Monte Carlo (MCMC) methods, a simulation-based estimation procedure. GLMMs introduce random effects into the linear predictors of generalized linear models (GLMs). Uninformative priors with normal distributions and wide variances were used to assess fixed effects, and inverse-Wishart distributions were specified for random effects. Comparisons of mean deviance statistics and deviance information criteria (DIC), a Bayesian generalization of the Akaike Information Criterion (AIC) that is well suited for complex hierarchical models, suggested inclusion of a random slope in the count portion for marijuana use produced the best model fit (DIC = 2228.18), whereas a random slope within both sub-models for cigarette use resulted in the best model fit (DIC = 841.50). Covariates included age, sex, ethnicity, and socioeconomic status (SES; i.e., highest parental education). Dummy codes were computed for Hispanic, Caucasian, Asian, African American, and Other, with Caucasian as the reference group. Socioeconomic status was a composite of mother’s and father’s highest education, which were coded 1= 8th grade or less to 6 = Advanced degree. Mean SES (M = 4.48, SD = 1.57) indicates that mothers and fathers, on average, attended some college. Sex was coded such that female was the reference group.



The data collected demonstrated extremely non-normal distributions for past 30-day cigarett (S = 10.58, K = 145.32) and marijuana (S = 4.72, K = 24.80) use quantity. In addition, an excessive amount of zeros were present in the data with 97.62% of zero responses for cigarette use and 91.86% of zero responses for marijuana, supporting the use of zero-altered models for count data. Past 30-day cigarette use prevalence was 2.4%. Past 30-day marijuana use prevalence was 8.1%. Past 30-day prevalence rates reported here are lower than national rates reported by the 2014 Youth Risk Behavior Survey (YRBS) for 9th graders (cigarette = 10.2% and marijuana = 17.7%) (Centers for Disease Control and Prevention, 2014) and Los Angeles area estimates in the YRBS for cigarette and marijuana use in 9th – 12th graders (6.7% and 20%, respectively; data on 9th graders only in Los Angeles not publically available). Since, nationally, from 9th to 12th grade, past 30-day prevalence of cigarette use increases 5.5% and marijuana use increases 5.7%, the data appear comparable with “Los Angeles School District” norms reported by the YRBS. Those reporting cigarette use, smoked, on average, one cigarette per day on the days that they used. Individuals reporting marijuana use smoked, on average, twice per day on the days that they used. Pearson product moment correlations among study variables are provided in Table 1. The significant correlation between inhibitory control and past 30-day amount of use should be interpreted with caution as the amount of use violates assumptions of normality and confounds both status and quantity, which is the rationale for employing the hurdle mixed-effects model approach.

Table 1
Means, Standard Deviations, and Correlations among Variables

Hurdle mixed-effects models

Table 2 depicts associations between inhibitory control and cigarette use prevalence and quantity of use. Among the covariates, SES was significantly and inversely associated with quantity of cigarette use. As hypothesized, inhibitory control was associated with 30-day prevalence of use such that greater inhibitory control was positively associated with membership in the certain zero group. A one unit increase in inhibitory control was associated with a 67% decrease in the odds of past 30-day cigarette use, and a two standard deviation increase in inhibitory control was associated with a 135% decreased odds of cigarette use. Inhibitory control was not significantly associated with quantity of use.

Table 2
Bayesian Mixed-effects Hurdle Model: Cigarettes

Table 3 depicts associations between inhibitory control and past 30-day marijuana use prevalence and quantity of use. Among the covariates, Asian-American adolescents were significantly less likely to have initiated use, compared to Caucasian adolescents. Asian-American students also reported significantly less quantity of marijuana use. SES was significantly associated with the likelihood of any marijuana use, a one-unit increase in amount of parental education predicting a 14% decrease in the odds of using marijuana on one or more occasions. Similarly, a one standard deviation increase in parental education was associated with a 29% decreased odds of using marijuana. Greater inhibitory control was significantly and positively associated with likelihood of membership in the certain zero category. A one unit increase in reported inhibitory control predicted a 42% decrease in the odds of marijuana use prevalence, and a two standard deviation increase was associated with a 80% decreased odds of marijuana use. Inhibitory control was not associated with quantity of marijuana use.

Table 3
Bayesian Mixed-effects Hurdle Model: Marijuana


Previous cross-sectional findings have demonstrated that both early and disordered use of substances have been associated with lower EF (Gruber, Dahlbren, Sagar, Gönenc, & Killgore, 2012; Harvey, Sellman, Porter, & Frampton, 2007), including inhibitory control (Gruber, Sagar, Dahlgren, Racine, & Lukas, 2012; Mathias, Blumenthal, Dawes, Liguori, Richard, Bray, et al., 2011). However, previous work has failed to disentangle prevalence from quantity of both cigarette and marijuana use. Consistent with study hypotheses, inhibitory control was significantly associated with a lower likelihood of ever using cigarettes and marijuana in the past 30 days. However inhibitory control was not significantly associated with amount of use. Thus, it appears that inhibitory control proficiency contributes to disrupting the impulse to initiate use of these two substances. However, when modeled together with past 30-day prevalence of use, inhibitory control’s relation to the amount of cigarettes and marijuana consumed was not significant.

Among the potential explanations for a lack of association between inhibitory control and amount of substance use is that once a substance is used, preexisting inhibitory control is disrupted and becomes a non-factor in the subsequent amount of use. Essentially, substances may level the neurocognitive playing field among those with proficient or poor inhibitory control when it comes to amount of use. A second potential reason for this lack of association is that the non-clinical sample consisted of adolescents who, in the vast majority of cases, were early in their progression to regular use. Thus, it is unlikely that many of the participants have progressed to amounts of use that would lead to chronic alteration in prefrontal cortical functioning.

Adult samples and/or clinical samples of adolescents might provide sufficient variance to detect associations between inhibitory control and amount of cigarette and/or marijuana use. The broader construct of “effortful control” has been shown to predict progression to problematic cigarette and marijuana use among emerging adults (Piehler, Véronneau, & Dishion, 2013). Similarly, adults with poorer inhibitory control tend to be more dependent, more prone to relapse, and have greater substance use frequency and substance use related consequences (Balfour, Wriete, Benwell, & Birrell, 2000; Li & Sinha, 2008; Powell, Dawkins, West, Powell, & Pickering, 2010). Thus, in adult substance misusers, poorer inhibitory control appears to enhance vulnerability to losing control over the impulse to use, potentially due to the neurobiologically mediated addiction process.

Relatedly, significant associations between inhibitory control and amount of cigarette and marijuana use may only exist at the very extreme end of the substance use continuum. For example, it is possible that among daily smokers inhibitory control deficit is associated with heavy dependent smoking but not light daily smoking (e.g., 10+ cigarettes per day vs. one cigarette per day). Since very few participants in this sample had progressed to extreme levels of substance use, it was not possible to explore this potential association. In short, the current sample may not have had enough variance across the entire continuum of substance use to detect significant associations between inhibitory control and amount of use. Future studies should explore whether current results generalize from this community sample of 9th graders to adolescents demonstrating disordered patterns of substance use.

Future research should also test potential mechanisms for the apparent disassociation in the relationships between inhibitory control and substance use prevalence and quantity of use. Among them are the psychoactive effects of nicotine and Δ9-tetrahydrocannabinol (THC) on the brain upon initiation of use (Goldstein & Volkow, 2011). For example, although nicotinic acetylcholine receptors are expressed throughout the brain, they are more densely expressed in the mesolimbic dopaminergic system of the brain (i.e., ventral tegmental area), responsible for reward sensitivity and motivation (De Biasi & Dani, 2011). Once used, the rewarding properties of nicotine on “bottom-up” motivation centers of the brain may make it difficult for adolescents, regardless of preexisting inhibitory control proficiency, to inhibit further use.

With respect to marijuana use, cannabinoid receptors, those receptors responsible for THC uptake, are highly expressed throughout the brain, including areas of the PFC associated with inhibitory control (Glass, Faull & Dragunow, 1997). Thus, in addition to enhanced “bottom-up” motivation to use, once marijuana is used, the direct acute effects of THC on the “top-down” regulatory processes governed by the PFC may become dysregulated. Clearly, although a growing body of research has begun to elucidate the complex and dynamic relationship between reward centers of the brain and the PFC as they relate to addiction (Goldstein & Volkow, 2011), much remains to be determined with respect to how these areas of the brain relate to early substance use as a predictor of substance use disorders.

Study findings should be considered in light of study limitations. The first is the reliance on self-reported inhibitory control. Various methods for measuring inhibitory control possess relative strengths and challenges. For example, objective behavioral measures, such as computerized tasks, are not subject to response biases or difficulty remembering and reporting inhibitory control-like behavior. Yet, such measures are time- and resource-intensive, which may be prohibitive within the context of large-scale studies in which assessments take place during limited school time, as was the case here (Riggs, 2015). Our previous research has demonstrated that youth can reliably self-report inhibitory control as early as the 4th grade (Riggs et al., 2012). However, it is possible that the EATQ-R is insensitive to the types of EF processes that are responsible for driving the progression from light user to heavy user.

The cross-sectional design precludes causal inference from findings. Although the data are cross-sectional, they were collected during a developmental period where those who use are considered early initiators. Since early prevalence is highly predictive of later misuse (Hawkins, Graham, Maguin, Abbott, Hill, & Catalano, 1997), a more complete understanding of the factors associated with early prevalence is of research importance. Compared to the more commonly used zero-inflated Poisson and negative binomial models, hurdle models are uniquely designed to assess differences in substance use among adolescents who have already used as well as differences in substance use between those who have use over the past 30 days and those who have not. Such comparisons were necessary to adequately test the hypotheses of the current study. However, longitudinal and experimental studies are needed to determine the causal and temporal parameters of the relation between inhibitory control and substance use.

Finally, the internal consistency of the 5-item EATQ-R was low. It is unclear why this may have been the case in this sample as it was similarly low across ethnicities and genders. However, consistent results were found between the 5-item EATQ-R and the more reliable 3-item EATQ-R, suggesting that the low reliability of the 5-item scale may not have dramatically influenced study findings.

It should also be noted that only when taking a social-ecological approach that views brain development as embedded within, and reciprocally influenced by, important family, peer, and educational contexts can a comprehensive understanding of brain development and its association with substance use be accomplished. Previous research has demonstrated brain by environment interactions with respect to nicotine use (Riggs & Pentz, in press), social and behavioral development (Blair, 2002), and trajectories of anti-social behavior (Moffitt, 1993). The current study provides an initial look at the association between inhibitory control and substance use, and future studies should explore potential moderating and mediating influences of social contexts in these relations.

One implication of the current results is that substance use preventive interventions that provide youth opportunities to practice self-regulation and promote inhibitory control may be more beneficial for promoting abstinence than preventing escalation among users. There currently exist a select number of school-based (i.e., Pre-K – 6th grade) preventive interventions which are evidence-based strategies for promoting inhibitory control as a mediator to early behavior problems (Bierman, Nix, Greenberg, Blair, & Domitrovich, 2008; Riggs, Greenberg, Kusché, & Pentz, 2006). Determining whether collateral prevention effects from these programs extend to the prevention of substance use would support the potential for the future application of current findings to substance use prevention.


1Exploratory factor analysis with a Varimax rotation was employed to examine the psychometric properties of the scale, and indicated the two reverse coded items yielded strong, positive loadings (i.e., .80 and .82) on a separate factor from the remaining items. One item demonstrated a negative association with the first factor, suggesting removal was appropriate (Knafl & Grey, 2007), but the second item provided a positive loading, although weak (i.e., .06). Results for the 3-item and 5-item scales were consistent. Therefore, results from the more reliable 3-item scale are presented.


1. Atkins DC, Baldwin SA, Zheng C, Gallop RJ, Neighbors C. A tutorial on count regression and zero-altered count models for longitudinal substance use data. Psychology of Addictive Behaviors. 2013;27:166–177. [PMC free article] [PubMed]
2. Balfour DJK, Wright AE, Benwell MEM, Birrell CE. The putative role of extra-synaptic mesolimbic dopamine in the neurobiology of nicotine dependence. Behavioral Brain Research. 2000;113:73–83. [PubMed]
3. Bierman KL, Nix RL, Greenberg MT, Blair C, Domitrovich CE. Executive functions and school readiness intervention: Impact, moderation, and mediation in the Head Start REDI program. Developmental Psychopathology. 2008;20:821–843. [PMC free article] [PubMed]
4. Blair C. School readiness: integrating cognition and emotion in a neurobiological conceptualization of children’s functioning at school entry. American Psychologist. 2002;57:111–127. [PubMed]
5. Bolla KI, Brown K, Eldreth D, Tate K, Cadet JL. Dose-related neurocognitive effects of marijuana use. Neurology. 2002;59:1337–1343. [PubMed]
6. Capaldi DM, Rothbart MK. Development and validation of an early adolescent temperament measure. Journal of Early Adolescence. 1992;12:153–173.
7. Centers for Disease Control and Prevention [Accessed on April 7, 2015];1991-2013 High School Youth Risk Behavior Survey Data. 2014 Available at
8. Chambers RA, Taylor JR, Potenza MN. Developmental neurocircuitry of motivation in adolescence: a critical period of addiction vulnerability. American Journal of Psychiatry. 2003;160:1041–1052. [PMC free article] [PubMed]
9. Davidson MC, Amso D, Anderson LC, Diamond A. Development of cognitive control and executive functions from 4 to 13 years: evidence from manipulations of memory, inhibition, and task switching. Neuropsychologia. 2006;44:2037–2078. [PMC free article] [PubMed]
10. De Biasi M, Dani JA. Reward, addiction, withdrawal to nicotine. Annual Review of Neuroscience. 2011;34:105–130. [PMC free article] [PubMed]
11. Dom G, Sabbe B, Hulstijn W, van den Brink W. Substance use disorders and the orbitofrontal cortex: systematic review of behavioural decision-making and neuroimaging studies. The British Journal of Psychiatry. 2005;187:209–220. [PubMed]
12. Eaton DK, Kann L, Kinchen S, Shanklin S, Ross J, Hawkins J, Wechsler H. MMWR CDC Surveillance Summaries. SS-5. 2010/06/04 ed Vol. 59. Atlanta, GA: 2010. Youth risk behavior surveillance - United States, 2009. [PubMed]
13. Ellis LK, Rothbart MK. Revision of the Early Adolescent Temperament Questionnaire. Poster presented at the 2001 Biennial Metting of the Society for Research in Child Development; Minneapolis, Minnesota. 2001.
14. Glass M, Faull RLM, Dragunow M. Cannabinoid receptors in the human brain: a detailed anatomical and quantitative autoradiographic study in the fetal, neonatal and adult human brain. Neuroscience. 1997;77:299–318. [PubMed]
15. Goldstein RZ, Volkow ND. Drug addiction and its underlying neurobiological basis: neuroimaging evidence for the involvement of the frontal cortex. The American Journal of Psychiatry. 2002;159:1642–1652. [PMC free article] [PubMed]
16. Goldstein RZ, Volkow ND. Dysfunction of the prefrontal cortex in addiction: neuroimaging findings and clinical implications. Nature Reviews Neuroscience. 2011;12:652–669. [PMC free article] [PubMed]
17. Gonzalez R. Acute and non-acute effects of cannabis on brain functioning and neuropsychological performance. Neuropsychological Review. 2007;17:347–361. [PubMed]
18. Gruber SA, Dahlgren MK, Sagar KA, Gönenc A, Killgore WD. Age of onset of marijuana use impacts inhibitory processing. Neuroscience Letters. 2012;511:89–94. [PMC free article] [PubMed]
19. Gruber SA, Sagar KA, Dahlgren MK, Racine M, Lucas SE. Age of onset of marijuana use and executive function. Psychology of Addictive Behaviors. 2012;26:486–506. [PMC free article] [PubMed]
20. Harvey MA, Sellman JD, Porter RJ, Frampton CM. The relationship between non-acute adolescent marijuana use and cognition. Drug and Alcohol Review. 2007;26:309–319. [PubMed]
21. Hawkins JD, Graham JW, Maguin E, Abbott R, Hill KG, Catalano RF. Exploring the effects of age of alcohol use prevalence and psychosocial risk factors on subsequent alcohol misuse. Journal of Studies on Alcohol and Drugs. 1997;58:280–290. [PMC free article] [PubMed]
22. Iacono WG, Carlson SR, Taylor J, Elkins IJ, McGue M. Behavioral disinhibition and the development of substance-use disorders: findings from the Minnesota Twin Family Study. Development and Psychopathology. 1999;11:869–900. [PubMed]
23. Ivanov I, Schulz KP, London ED, Newcorn JH. Inhibitory control deficits in childhood and risk for substance use disorder: a review. The American Journal of Drug and Alcohol Abuse. 2008;34:239–258. [PubMed]
24. Johnston LD, O'Malley PM, Bachman JG, Schulenberg JE. Quantity I: Secondary school students. National Institute on Drug Abuse; Bethesda, MD: 2010. Monitoring the Future national survey results on drug use, 1975-2009; p. 305. NIH Publication No. 10-7584.
25. Knafl GJ, Grey M. Factor analysis evaluation through likelihood cross-validation. Statistical Methods in Medical Research. 2007;16:77–102. [PMC free article] [PubMed]
26. Li CR, Sinha R. Inhibitory control and emotional stress regulation: neuroimaging evidence for frontal-limbic dysfunction in psycho-stimulant addiction. Neuroscience & Biobehavioral Reviews. 2008;32:581–597. [PMC free article] [PubMed]
27. Mahone EM, Cirino PT, Cutting LE, Cerrone PM, Hagelthorn KM, Hiemenz JR, et al. Validity of the behavior rating inventory of executive funtion in children with ADHD and/or Tourette synrome. Archives of Clinical Neuropsychology. 2002;17:643–662. [PubMed]
28. Mathias CW, Blumenthal TD, Dawes MA, Liguori A, Richard DM, Bray B, et al. Failure to sustain prepulse inhibition in adlescent marijuana users. Drug and Alcohol Dependence. 2011;116:110–116. [PMC free article] [PubMed]
29. McCulloch CE, Searle SR, Neuhaus JM. Generalized linear mixed models. 2nd ed John Wiley & Sons, Inc; Hoboken, NJ: 2008.
30. Moeller FG, Barratt ES, Fischer CJ, Dougherty DM, Reilly EL, Mathias CW, et al. P300 event-related potential amplitude and impulsivity in cocaine-dependent subjects. Neuropsychobiology. 2004;50:167–173. [PubMed]
31. Moffitt TE. Adolescence-limited and life-course persistent antisocial behaviour: A developmental taxonomy. Psychological Review. 1993;100:674–701. [PubMed]
32. Munakata Y, Herd SA, Chatham CH, Depue BE, Banich MT, O’Reilly RC. A unified framework for inhibitory control. Trends in Cognitive Sciences. 2011;15:453–459. [PMC free article] [PubMed]
33. Muris P, Meesters C. Reactive and regulative temperament in youths: Psychometric evaluation of the Early Adolescent Temperament Questionnaire-Revised. Journal of Psychopathology and Behavioral Assessment. 2009;31:7–19.
34. Neighbors C, Atkins DC, Lewis MA, Lee CM, Kaysen D, Mittmann A, Rodriguez LM. Event specific drinking among college students. Psychology of Addictive Behaviors. 2011;25:702–707. [PMC free article] [PubMed]
35. Nigg JT, Wong MM, Martel MM, Jester JM, Puttler LI, Glass JM, et al. Poor response inhibition as a predictor of problem drinking and illicit drug use in adolescents at risk for alcoholism and other substance use disorders. Journal of the American Academy of Child & Adolescent Psychiatry. 2006;45:468–475. [PubMed]
36. Norman AL, Pulido C, Squeglia LM, Spadoni AD, Paulus MP, Tapert SF. Neural activation during inhibition predicts prevalence of substance use in adolescence. Drug and Alcohol Dependence. 2011;119:216–223. [PMC free article] [PubMed]
37. Piehler TF, Véronneau MH, Dishion TJ. Substance use progression from adolescence to early adulthood: effortful control in the context of friendship influence and early-onset use. Journal of Abnormal Child Psychology. 2012;40:1045–1058. [PMC free article] [PubMed]
38. Powell J, Dawkins L, West R, Powell J, Pickering A. Relapse to smoking during unaided cessation: clinical, cognitive and motivational predictors. Psychopharmacology. 2010;212:537–549. [PubMed]
39. Reyna VF, Farley F. Risk and rationality in adolescent decision making: Implications for theory, practice, and public policy. Psychological Science in the Public Interest. 2006;7:1–44. [PubMed]
40. de Ridder DTD, Lensvelt-Mulders G, Finkenauer C, Stok FM, Baumeister RF. Taking stock of self-control: a meta-analysis of how trait self-control relates to a wide range of behaviors. Personality and Social Psychology Review. 2012;16:76–99. [PubMed]
41. Riggs NR. Translating research in neurocognitive development to interventions for youth substance use. Current Addiction Reports. 2015;2:114–124. [PMC free article] [PubMed]
42. Riggs NR, Chou C-P, Li C, Pentz MA. Adolescent to emerging adulthood smoking trajectories: When do smoking trajectories diverge and do they predict early adulthood nicotine dependence? Nicotine & Tobacco Research. 2007;9:1147–1154. [PubMed]
43. Riggs NR, Greenberg MT, Kusché CA, Pentz MA. The mediational role of neurocognition in the behavioral outcomes of a social-emotional prevention program in elementary school students: Effects of the PATHS Curriculum. Prevention Science. 2006;7:91–102. [PubMed]
44. Riggs NR, Pentz MA. Inhibitory control and the onset of combustible cigarette, e-cigarette, and hookah use in early adolescence: the moderating role of socioeconomic status. Child Neuropsychology. (in press) [PubMed]
45. Riggs NR, Spruijt-Metz D, Chou CP, Pentz MA. Relationships between executive cognitive function and lifetime substance use and obesity-related behaviors in fourth grade youth. Child Neuropsychology. 2012;18:1–11. [PubMed]
46. Riggs NR, Tate EB, Ridenour TA, Reynolds MD, Zhai ZW, Vanyukov MM, et al. Longitudinal associations from neurobehavioral disinhibition to adolescent risky sexual behavior in boys: direct and mediated effects through moderate alcohol consumption. Journal of Adolescent Health. 2013;53:465–470. [PMC free article] [PubMed]
47. Rothbart MK. Temperament in childhood: A framework. In: Kohnstamm GA, Bates JA, Rothbart MK, editors. Temperament in childhood. Wiley; New York: 1989. pp. 59–73.
48. Stein EA, Pankiewicz J, Harsch HH, Cho J-K, Fuller SA, Hoffmann RG, Hawkins M, Rao SM, Bandettini PA, Bloom AS. Nicotine-induced limbic cortical activation in the human brain: a functional MRI study. The American Journal of Psychiatry. 1998;155:1009–1015. [PubMed]
49. Steinberg L, Dahl R, Keating D, Kupfer DJ, Masten AS, Pine D. The study of developmental psychopathology in adolescence: Integrating affective neuroscience with the study of context. In: Cicchetti D, Cohen D, editors. Handbook of developmental psychopathology. 2nd ed Wiley; Hoboken, NJ: 2006. pp. 710–741.
50. Sussman S, Dent C, Stacy A. Project Towards No Drug Abuse: A review of the findings and future directions. American Journal of Health Behavior. 2002;26:354–365. [PubMed]
51. Tarter RE, Kirisci L, Mezzich A, Cornelius JR, Pajer K, Vanyukov M, et al. Neurobehavioral disinhibition in childhood predicts early age at onset of substance use disorder. American Journal of Psychiatry. 2003;160:1078–1085. [PubMed]
52. Trucco EM, Colder CR, Wieczorek WF. Vulnerability to peer influence: a moderated mediation study of early adolescent alcohol use prevalence. Addictive Behaviors. 2011;36:729–736. [PMC free article] [PubMed]
53. Zelazo PD, Carlson SM, Kesek A. The development of executive function in childhood. In: Nelson C, Luciana M, editors. Handbook of Developmental Cognitive Neuroscience. 2nd Ed MIT Press; Cambridge, MA: 2008. pp. 553–574.