Citation
Predictors of Daily Marijuana Use among a National Sample of US Youth

Material Information

Title:
Predictors of Daily Marijuana Use among a National Sample of US Youth
Creator:
Khan, Shivani Rajeshwori
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Epidemiology
Committee Chair:
STRILEY,CATHERINE L
Committee Co-Chair:
COTTLER,LINDA B
Committee Members:
WINTERSTEIN,ALMUT GERTRUD
NIXON,SARA J
Graduation Date:
12/17/2017

Subjects

Subjects / Keywords:
adolescents
marijuana
prevention

Notes

General Note:
Marijuana is legal for medical and recreational purposes in an increasing number of states in the United States. With this legalization, an increased availability of marijuana and a decreased perception of harm from its regular use have occurred especially among adolescents, both of which may increase the likelihood of regular marijuana use. Regular marijuana use, commonly defined as daily use, during adolescence has been associated with an increased vulnerability to neurobiological, psychosocial and other adverse health outcomes. To prevent these effects, it is important to understand factors that influence current daily marijuana use among youth who have already initiated marijuana use. Recognizing this need, this dissertation aimed to analyze, among youth who had initiated marijuana use: 1) the individual-level factors, 2) the sex- and racial/ethnicity-specific individual-level factors, and 3) the relationship-level factors, both direct and indirect effects, that influence current daily marijuana use. Data came from the National Monitoring of Adolescent Prescription Stimulants Study (N-MAPSS; PI Linda B. Cottler), a cross-sectional study conducted from 2008 to 2011 which examined a broad range of risk factors for prescription stimulant use and misuse among 11,048 youth aged 10 to 18 years from 10 US metropolitan areas using an entertainment-venue intercept method. Our analyses included youth from 13 to 18 years of age. Sociodemographic, mental health and behavioral factors (lifetime and current use of substances such as tobacco, alcohol, marijuana and other drugs), and parental and peer-related variables were assessed and included in these analyses. The main outcome of the study was current daily marijuana use, defined as use on 20 or more days in the past 30 days. Current daily marijuana use was compared with no current marijuana use (Chapters 3 and 4) and with current non-daily marijuana use (Chapter 5). This dissertation provides an in-depth understanding of the influence of individual- and relationship-level factors on current daily marijuana use among youth who initiated marijuana use. Findings can be used to help target prevention efforts to the specific risk factors for daily marijuana use among teens.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
12/31/2019

Downloads

This item has the following downloads:


Full Text

PAGE 1

PREDICTORS OF DAILY MARIJUANA USE AMONG A NATIONAL SAMPLE OF US YOUTH By SHIVANI RAJESHWORI KHAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR T HE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2017

PAGE 2

2017 Shivani Rajeshwori Khan

PAGE 3

Dedicated to Mummy, Buwa, Babu, Fuwa and Jijumuwa

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my sincere gratit ude to my advisor, Dr. Catherine W. Striley, for her excellent mentorship and unfaltering support throughout my Ph D program. I have gained immense knowledge and experience in my research and writing from her. I would also like to thank my committee membe rs, Dr. Linda B. Cottler, Dr. Al mut G. Winterstein and Dr. Sara J. Nixon for their expert guidance and invaluable feedback. Diana Rojas, Johanny Perez, Mirsada Serdarevic, Vicki Osbo rne, Sadaf Milani, Sunil Swami, Hannah Crooke, Rosha Poudel, Steve Chrzanowski, Denny Fe Agana and Pooja Subedi. I cherish the moments of friendship and laughter spent with them and look forward to many such moments in the future. I am also very thankful to my cousin, Shubhekchya Malla, and my best friends, Shree Baba Pokharel, Shringkhala Bajimaya, Karuna Giri and Abhishek Chan Tiwari, who are very close to my heart. Each of them has provided me wi th constant encouragement and support at all phases of my life and my life would not ha ve been the same without them. I would also like to thank my family for always being there for me. Lastly, I would like to thank the Department of Epidemiology at UF and al l the participants of N MAPSS. Without their contribution, my dissertation would not have been possible.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FI GURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ............................. 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 1.1 Ep idemiology of Youth Marijuana Use ................................ .............................. 12 1.2 Definition of Daily Marijuana Use ................................ ................................ ...... 15 1.3 Conceptual Framework ................................ ................................ ..................... 15 2 DATA SOURCE ................................ ................................ ................................ ...... 19 2.1 National Monitoring of Adolescent Prescription Stimulants Study (N MAPSS) ................................ ................................ ................................ ............... 19 2.2 Assessment ................................ ................................ ................................ ...... 20 3 INDIVIDUAL LEVEL FACTORS ASSOCIATED WITH CURRENT DAILY MARIJUANA USE AMONG US YOUTH WHO HAD INITIATED MARIJUANA USE ................................ ................................ ................................ ........................ 25 3.1 Background ................................ ................................ ................................ ....... 25 3.2 Methods ................................ ................................ ................................ ............ 29 3.2.1 Sample ................................ ................................ ................................ .... 30 3.2.2 Measures ................................ ................................ ................................ 30 3.2.2.1 Current marijuana use ................................ ................................ .... 30 3.2.2.2 Sociodemographic factors ................................ .............................. 31 3.2.2.3 Behavioral factors ................................ ................................ .......... 31 3.2.2.4 Mental health factors ................................ ................................ ...... 32 3.3 Statistical Analyses ................................ ................................ ........................... 32 3.4 Results ................................ ................................ ................................ .............. 33 3.5 Discussion ................................ ................................ ................................ ........ 36 3.6 Conclusion ................................ ................................ ................................ ........ 40 4 SEX AND RACE/ETHNICITY SPECIFIC INDIVIDUAL LEVEL FACTORS ASSOCIATED WITH CURRENT DAILY MARIJUANA USE AMONG YOUTH ....... 44 4.1 Background ................................ ................................ ................................ ....... 44 4.2 Methods ................................ ................................ ................................ ............ 49 4.2.1 Sample ................................ ................................ ................................ .... 50

PAGE 6

6 4.2.2 Measures ................................ ................................ ................................ 51 4.2.2.1 Current marijuana use ................................ ................................ .......... 51 4.2.2.2 Individual level factors ................................ ................................ .......... 51 4.3 Statistical Ana lyses ................................ ................................ ........................... 53 4.4 Results ................................ ................................ ................................ .............. 53 4.4.1 Sex Differences ................................ ................................ ....................... 53 4.4.2 Racial/Et hnic Differences ................................ ................................ ........ 55 4.5 Discussion ................................ ................................ ................................ ........ 57 4.6 Conclusion ................................ ................................ ................................ ........ 63 5 INFLUE NCE OF PARENTAL AND PEER FACTORS ON CURRENT DAILY MARIJUANA USE AMONG US YOUTH WHO HAD INITIATED MARIJUANA USE ................................ ................................ ................................ ........................ 70 5.1 Background ................................ ................................ ................................ ....... 70 5.2 Methods ................................ ................................ ................................ ............ 74 5.2.1 Sample ................................ ................................ ................................ .... 74 5.2.2 Measures ................................ ................................ ................................ 74 5.3 Statist ical Analyses ................................ ................................ ........................... 77 5.4 Results ................................ ................................ ................................ .............. 78 5.4.1 Direct Effects ................................ ................................ ........................... 79 5.4.2 In direct Effects ................................ ................................ ......................... 79 5.5 Discussion ................................ ................................ ................................ ........ 80 5.6 Conclusion ................................ ................................ ................................ ........ 85 6 CONCLUSI ON ................................ ................................ ................................ ........ 90 6.1 Main Findings ................................ ................................ ................................ ... 91 6.2 Strengths and Limitations ................................ ................................ ................. 96 6.3 Future Research ................................ ................................ ............................... 97 APPENDIX: NATIONAL MONITORING OF ADOLESCENT PRESCRIPTION STIMULANTS STUDY ................................ ................................ ............................ 99 LIST OF REFERENCES ................................ ................................ ............................. 102 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 113

PAGE 7

7 LIST OF TABLES Table page 1 1 Terms and definitions used for daily mar ijuana use in different studies .............. 18 3 1 Sociodemographic, behavioral and mental health characteristics by current marijuana use among youth 13 to 18 years of age who had initiated marijuana use in N MAPSS (n=3,100) ................................ ................................ 42 3 2 Multinomial logistic regression model of sociodemographic, behavioral and mental health characteristics for current marijuana use among youth 13 to 18 years of age who had initiated marijuana use in N MAPSS (n=3,100) ............... 43 4 1 Sociodemographic, behavioral and mental health characteristics by current daily marijuana use among male and female youth 13 to 18 years of age who had initiated marijuana use in N MAPSS ................................ ............................ 64 4 3 Sociodemographic, behavioral and mental health characteristics and current daily marijuana use among NH White and NH Black youth 13 to 18 years of age who had initiated marijuana use in N MAPSS ................................ ............. 66 4 4 Sociodemographic, behavioral and mental health characteristics and current daily marijuana use among Hispanic youth 13 t o 18 years of age who had initiated marijuana use in N MAPSS ................................ ................................ ... 67 4 5 Multinomial logistic regression models of sociodemographic, behavioral and mental health characteristics for current marijua na use among NH White and NH Black youth 13 to 18 years of age who had initiated marijuana use in N MAPSS ................................ ................................ ................................ ............... 68 4 6 Multinomial logistic regression model of sociodemographic, behavioral and menta l health characteristics for current marijuana use among Hispanic youth 13 to 18 years of age who had initiated marijuana use in N MAPSS ........ 69 5 1 Characteristics of youth 13 to 18 years of age w ho had initiated marijuana use overall and by current daily and current non daily marijuana use in N MAPSS Waves Three and Four (n=1,573) ................................ ......................... 87 5 2 Correlation coefficients between variables incl uded in the model ....................... 88 A 1 N MAPSS questions used in analyses ................................ ............................... 99

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Socio ecological model ................................ ................................ ...................... 18 2 1 N MAPSS recruitment sites map ................................ ................................ ....... 23 2 2 Flow chart for outcome measures for Chapters 3 and 4 ................................ .... 23 2 3 Flow chart for outcome measures for Chapter 5 ................................ ................ 24

PAGE 9

9 LIST OF ABBREVIATIONS ADD Attention Deficit Disorder ADHD Attention Deficit Hyperactivity Disorder AGFI Adjusted Goodness of Fit Index AOR Adjusted Odds Ratio CFI Comparative Fit Index CI Confidence Interval DAWN Drug Abuse and Warning Network GFI Goodness of Fit Index MTF Monitoring the Future NH Non Hispanic N MAPSS National Monitoring of Ad olescent Prescription Stimulant s Study NSDUH National Survey on Drug Use and Health RBA Risk Behavior Assessment REF Reference Group RMSEA Root Mean Square Error of Approximation SAM Substance Abuse Module SAMHSA Substance Abuse and Mental Health Adm inistration SES Socio economic status SAM Substance Abuse Module THC Tetrahydrocannabinol VIF Variance Inflation factor YRBS Youth Risk Behavior Survey

PAGE 10

10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PREDICTORS OF DAILY MARIJUANA USE AMONG A NATIONAL SAMPLE OF US YOUTH By Shivani Rajeshwori Khan December 2017 Chair: Catherine W. Striley Major: Epidemiology Marijuana is legal fo r medical and recreational purposes in an increasing number of states in the United States. With this legalization, an increased availability of marijuana and a decreased perception of harm from its regular use have occurred especially among adolescents, b oth of which may increase the likelihood of regular mariju ana use. Regular marijuana use, commonly defined as daily use during adolescence has been associated with an increased vulnerability to neurobiological, psychosocial and other adverse health outcom es. To prevent these effects, it is important to understand factors that influence current daily marijuana use among youth who have already initiated marijuana use. Recognizing this need, t his dissertation aimed to analyze among youth who had initiated ma rijuana use: 1) the individual level factors 2) the sex and racial/ethnicity specific individual level factors and 3) the relationship level factors both direct and indirect effects that influence current daily marijuana use. Data came from the Natio nal Monitoring of Adolescent Prescription Stimulants Study (N MAPSS; PI Linda B. Cottler), a cross sectional study conducted from 2008 to 2011 which examined a broad range of risk factors for prescription stimulant use and

PAGE 11

11 misuse among 11,048 youth aged 10 to 18 years from 10 US metropolitan areas using an entertainment venue int ercept method. Our analyses included youth from 13 to 18 years of age. S ociodemographic, mental health and behavioral factors ( lifetime and current use of substances such as tobacco alcohol, marijuana and other drugs ) and parental and peer related variables were assessed and included in these analyses The main outcome of the study was current daily marijuana use, defined as use on 20 or more days in the past 30 days. Current daily marijuana use was compared w ith no current marijuana use (Chapters 3 and 4) and with current non daily marijuana use (Chapter 5). This dissertation provides an in depth understanding of the i nfluence of individual and relationship level factors on curre nt daily marijuana use among youth who i nitiated marijuana use Findings can be used to help target prevention efforts to the specific risk factors for daily marijuana use among teens.

PAGE 12

12 CHAPTER 1 INTRODUCTION 1.1 Epidemiology of Youth Marijuana Use Mariju ana is the most commonly used federally prohibited drug in the United States The 2016 National Survey on Drug Use and Health (NSDUH) reported that 48.5 % of the US population (12 years or older) endorsed lifetime marijuana use. Further, the past month prev alence s of ma rijuana use were reported to be 6.5%, 20.8% and 7.2 % among the age groups 12 to 17, 18 to 25 and 26 and o lder, respectively in 2016 (SAMHSA, 2017 ). Although the rates o f use are higher among 18 to 25 y ear olds compared to other age groups, mar ijuana use among ado lescents is a particular pu blic health concern because brain development continues throughout adolescence and into early adulthood (Gogtay et al., 2004) Durin g these developmenta l stages, the brain is more susceptible than a mature brain to adverse effects of substances such as marijuana that can alter brain structure and functioning (Dinieri and Hurd, 2012) Moreover, adolescence is the period of highest risk of substance initiation and experimentation. Nearly one half of high school seniors reported marijuana use at least once in their lifetime (Johnston et al., 2016). T he health effects associated with marijuana use have been widely debated, though adverse effects associated with marijuana use during adolesce nce and adverse effects of regular use have been consistent in the literature (Az ofeifa, 2016) In epidemiology, most studies define regular marijuana use as daily or almost daily use (Hall and Degenhardt, 2009) Short term effects of marijuana use among y outh include impaired short term memory that diminishes the ability to learn and retain information, reduced motor coordination that increases the risk of motor vehicle accidents and injuries,

PAGE 13

13 altered judgment increasing the likelihood of risky behaviors paranoia and psychosis when used in higher doses (Volkow et al., 2014) The Drug Abuse Warning Network (DAWN) reported an increase in emergency department (ED) visits in which individuals indicated marijuana use in the medical record; 13% of such visits were made by youth between 12 and 17 years of age (DAWN, 2013). Moreover, the long term or regula r use of marijuana may cause more detrimental effects such as cannabis use disorder. About 17% of those who begin use in adolescence and about 25 50% of those who use marijuana regula rly become addicted to the drug (Hall and Degenhardt, 2009) Other identified effects of long term or regular use range from poor educational outcomes, altered brain development, cognitive impairment, respiratory problems (e.g. chronic bronchitis), diminished satisfaction in life achievement to risk of mental problems such as schizophrenia, particularly among those with a predisposition to such disorders (Volkow et al., 2014) Currently, twenty nine states and the District of Columbia have passed some form of marijuana legalization; eight states and the District of Columbia have passed mariju ana legalization laws for recreational purposes (Kilmer and MacCoun, 2017). With the increase in marijuana legalization in a growing number of states, there is a potential for increase in availability of marijuana. Recent years have also shown a decrease i n perception of harm of regular use among adolescents (Johnston et al., 2016). The proportion of high school seniors who perceived regular marijuana use as harmful declined from 58.3% in 2000 to 31% in 2016. The prevalence of daily marijuana use, defined b y the Monitoring the Future (MTF) study as use of marijuana on 20 or more occasions in the past 30 days, escalated by grade level with 0.7%, 2.5% and 6% of 8 th

PAGE 14

14 10 th and 12 th g rade students reporting daily marijuana use, respectively in 2016. Further, dail y marijuana use exceeded daily tobacco use among 10 th (2.5% vs 1.9%) and 12 th (6.0 vs 4.8%) grade students (Johnston et al., 2016). Not all youth who use marijuana progress to daily marijuana use. Research indicates that about 10% of those who initiate mar ijuana use become daily users (H all and Degenhardt, 2014) Also, a majority of research has focused on the risk and protective factors of youth marijuana use by comparing lifetime or past 30 day use versus no use (SAMHSA, 2014) and neglected to make a distinctio n among marijuana users. T he assumption that marijuana using populations are homogeneous can be misleading; there are different types of users depending on the frequency of use For instance, among marijuana using college students, daily or nearly daily users were different from t hose who used a few times in a month with daily users having more risk factors and consequences compared to the other groups (Pearson et al., 2017). Our review indicated that the major p ublic health concern lie s in the adverse consequences associated wit h reg ular use (daily) more than with lifetime or current use. The NSDUH definition for daily marijuana use, use on 20 or more days in the last 30 days, is a conservative measure of daily marijuana use. This definition has been used for reporting daily mari juana use for several years in the literature By using this definition of current daily marijuana use, we could identify risk factors specific to such use. The risk factors identified could then be incorporated in prevention programs to reduce the public health consequences associated with such use. Therefore, this dissertation intended to identify individual and relationship level factors specific to current daily marijuana use

PAGE 15

15 among youth who had initiated marijuana use usi ng a national sample of youth and the NSDUH definition 1.2 Definition of Daily Marijuana Use Studies differ in their definition of daily marijuana use. For instance, the annual cross sectional national surveys on adolescent substance use, the National Survey on Drug Use and Health (N SDUH) and the Monitoring the Future (MTF), differ in t heir definitions of daily use. The NSDUH defines past month daily or almost daily marijuana use as use on 20 or more days in the past 30 days; the MTF defines daily use as use on 20 or more occasions in the past 30 days. According to the MTF definition, 20 or more occasions can occur over a weekend or it may possibly occur as one occasion per day. Other studies with small sample sizes defined daily use as having used every day or used on at least 6 days in a week (Conroy et al., 2016; Silins et al., 2014) Further, studies have also used different terms such as use, frequent use or regular use with varying definitions and hav e focused on specific risk factors (Tu et al., 2008; Griffin et al., 2002; Creemers et al., 2010; van der Pol et al., 2013; Farhat et al., 2011). Tu and colleagues (2008) defined heavy use as having used 10 or more times in the past 30 days; Griffin and co lleagues (2002) defined regular use as having used on a monthl y basis or more frequently; van der Pol and colleagues (2013) defined frequent use as having used on at least three days in a week for at least 12 months. Given this landscape, it is difficult t o compare and replicate findings from different studies. Thus, risk factors specific to daily marijuana use remain unclear. 1.3 Conceptual Framework Research indicated that s ocial learning theory has been widely applied to explain deviant behavior includi ng substance use among adolescents. Along with the individual

PAGE 16

16 level factors that have been found to be associated with marijuana use among youth, social learning theory focus es on relationship level factors, such as parental and peer influence on youth mar ijuana use. According to the social learning theory, children initially learn social norms and skills from their parents but as they reach adolescence, they begin to model the norms and skills from their peers As youth mature, their associations with peer s become more important and peers become more influential than parents to their behavior If they associate with deviant pe ers this may cause additional detachment from parents increasing the likelihood of their involvement in deviant behavior such as sub stance use (Akers et al., 1979). Parental variables that are associated with marijuana use among youth include family disruption (Low et al., 2012), parental tobacco or marijuana use (Miller et al., 2013; Swift et al., 2008), poor parental monitoring (Deve r et al., 2012; White et al., 2006), low parental attachment (Heavyrunner et al., 2010), and living with divorced or separated parents (Mandara et al., 2011). Peer variables associated with youth marijuana use include peer marijuana use (Ali et al., 2011; Pinchevsky et al., 2012), having substance using peers (Korhonen et al., 2008; Perkonigg et al., 2008; Mrug et al., 2010) and affiliation with deviant peers (Hampson et al., 2008). I ndivi dual level factors that are associated with youth marijuana use inc lude poor academic performance (Tu et al., 2008) anxiety (Marmorstein et al., 2010) depression (Pacek et al., 2012; Tu et al., 2008) conduct disorder (Wymbs et al., 2012) aggres sive behavior (Farhat et al., 2011; Korhonen et al., 2008) attention problems (King et al., 2011), impulsivity (Heavyrunner Rioux and Hollist, 2 010), antisocial personality disorder (Perkonigg et al., 2008), tobacco use (Guxens et al., 2007; Korhonen et al., 2008; Tu et

PAGE 17

17 Perkonigg et al., 2008; Tu et al., 2008), lack of extracurricular activities (Perez et al., 2011) and positive attitudes and perception towards marijuana use (Hayaki et al., 2010; King et al., 2012; Morrison et al., 2010; Stephens et al., 2009). Early onset of marijuana use has also been linked w ith re gular marijuana use among youth (Terry McElrath and ; Sagar et al., 2015 ). Another widely applied model in substance use is the socio ecological model. In 2014, the Substance Abuse and Mental Health Services Administration (SAMHSA) Cent er for the Application of Prevention Technologies (2014) listed important risk and protective factors assoc iated with youth marijuana use utilizing a socio ecological framework That framework included four levels of factors: individual, relationship, comm unity and societal. Community level factors (neighborhoods, work places, schools and institutions) and s ocietal level factors (culture, norms and access, historical trauma, media, and laws and policies) are also very important in predicting substance use. However, the influences of these factors are more distal compared to individual and relationship level factors own characteristics and the characteristics of their importa nt relationships, this dissertation focused only on individual and relationship level factors that may affe ct current marijuana use

PAGE 18

18 Figure 1 1 Socio ecological model Table 1 1 Terms and definitions used for daily marijuana use in different s tudies Study Term used Definition NSDUH Daily or almost daily use Use on 20 or more days in the past 30 days MTF Daily use Use on 20 or more occasions in the past 30 days Silins et al., 2014 Daily use Use everyday Conroy et al., 2016 Daily use At least 6 days in a week Tu et al., 2008 Heavy use Use 10 or more times in the past 30 days Griffin et al., 2002 Regular use Use on a monthly basis or more frequently Creemers et al., 2010 Regular use Use on at least 4 occasions in the past four weeks Johnson et al., 2015 Frequent use Use 20 to 39 times in the past 30 days van der Pol et al., 2013 Frequent use Use on at least 3 days per week for at least 12 months Farhat et al., 2011 Current frequent use Use at least 3 times in the past 30 days

PAGE 19

19 CHAPTER 2 DATA SOURCE 2.1 National Monitoring of Adolescent Prescription Stimulants Study (N MAPSS) Data used in the three papers herein came from the National Monitoring of Adolescent Prescription Stimulants Study (N MAPSS) (PI Dr. Linda B. Cottler). N MAPSS was c onducted under the contract with Pinney Associates, Inc., with the funding provided by Shire Development LLC and Noven Pharmaceuticals. The main objective of this study was to investigate the prevalence of and risk factors for misuse, abuse and diversion of prescription stimulants am ong youth. N MAPSS was a cross sectional study that was conducted in urban, suburban and rural areas of 10 US cities in four waves: Fall 2008, spring 2009, fall 2010 and spring 2011. These cities were selected on the basis of p rescription stimulant use patterns. The IMS Health database contains records of prescriptions dis pensed at each retail pharmacy in the US. Using the 2008 database, states with the highest volume of stimulant prescriptions were identified and the cities wit hin those states with the highest volume of stimulant prescriptions were selected. Four cities from the eastern US (Boston, MA; New York, NY; Philadelphia, PA and Tampa Florida) three cities from the central US (St Louis, MI; Cincinnati, OH and H ou ston, TX) and three cities from the western US (Denver, CO; Los Angeles, CA and Seattle, Washington) were included in the study (Figure 2 1). Recruiters/trained interviewers approached 21,444 youth for participation at shopping malls, movie theatres, arcades l ibraries, skate parks, parks and sports and recreational centers using an entertainment venue intercept met hod. This method allowed the capturing of a large and diverse sample of youth by recruiting individuals from popular venues where they often visit or gather. Eligibility criteria included youth

PAGE 20

20 between 10 to 18 years of age residing in urban, suburban or rural areas in the selected cities. These areas were categorized based on city limits, proximity to city limits a nd population density. Eligibility cr iteria ex cluded those who were college students, non English readers, and those who were cognitively impaired or did not know of their zip code (Cottler et al., 2013). Among 21,444 youth, a quarter did not stop to listen to their introduction, 10% stopped but refused to participate in the study and 21% were ineligible due to their age or zip code. Out of 11,468 youth who participated, 3.7% were eliminated because of untruthful responses, duplicates, incompletes and ineligibility. A total of 11,048 youth 10 to 18 years of age was included in the final sample which represented a participation rate of 86.7% (Cotter et al., 2013). Trained interviewers obtained verbal consents from each participant for anonymous interviews and provi ded verbal guidance to 10 to 1 2 year olds through the survey whereas it was self administered amo ng 13 to 18 year olds. Participants took about 20 30 minutes to complete the survey and received a remuneration of a gift card worth $10 from a national electronics store upon completion. T he study protocol was approved by Washington University in St. Louis and the University of Florida Institutional Review Boards (Cottler et al., 2013; Lasopa et al., 2015; Wang et al., 2015; Striley et al., 2016). 2.2 Assessment The N MAPS S survey w as devel oped by the Epidemiology and Prevention Research Group (EPRG) led by Dr. Linda Cottler and her team at the Washington University School of M edicine (WUSM ). The survey questions were adapted from the Substance Abuse Module (SAM) (Horton et al., 2000) and t he Washington University Risk Behavior Assessment (RBA) (Shacham and Cottler, 2010) and it consisted of two

PAGE 21

21 parts. Along with questions that asked youth to identify the stimulants by name and formulation, Part O ne covered questions that assessed sociodemog raphic and mental health variables of youth. S ociodemographic variables that we used for this dissertation were age, sex, race/ethnicity, family structure and grades in school. Mental health variables that we used were the self reported diagnosis of ADD/AD HD, and symptoms of anxiety, conduct disorder and depression. After t he completion of Part O ne, Part T wo covered questions that assessed lifetime and current use of stimulants and opioids as well as other substances such as tobacco, alcohol and mariju an a Behavioral variables that were used in this dissertation included current tobacco use, current alcohol use, current marijuana use including current daily use, the age of marijuana ini tiation and lifetime other drug (cocaine or crack, heroin, club drugs l ike ecstasy, hallucinogens like LSD or mushrooms, anabolic steroids, methamphetamine or inhalants) use. Additionally, other questions that assessed parental warnings against substance use and peer su bstance use were added in Part T wo of Waves Three and F ou r. Parental warning questions used in this dissertation included how often parent s or guardian s had warned youth to not use marijuana or alcohol or to smoke or chew tobacco and if youth had ever gotten alcohol from their parent. Youth were also asked if an yone in their household smoked cigarettes or cigars or used chewing tobacco, snuff or dip. Peer substance use questions included tobacco and if they had ever gotten alc ohol from their friends. Current marijuana use status: The main outcome of interest in this study was current marijuana use (past 30 day use) among youth who have used marijuana. We

PAGE 22

22 excluded 10 to 12 year olds due to low rates of self reported marijuana u se. To assess the outcome variab le, all participants were asked if they had ever used marijuana. Only further asked responses for this question were categorized into zero days, one to two days, three to five days, six to nine days, ten to nineteen days, twenty to twenty nine days and all thirty in the last 30 nine days and all thirty 13 to 18 years of age who had initiated marijuana use from all four waves of N MAPSS (n=3,155) P arental and peer related vari abl es that I used for chapter 5 were only added in Waves Three and F our of N MAPSS. T herefore, chapter 5 included youth 13 to 18 years of age who had initiated marijuana use from last tw o waves (n=1,586 ). Overall, N MAPSS consisted of a large and diverse sample of youth from urban, suburban and rural areas across 10 US cities. It included a significant proportion of youth who endorsed marijuana use and provided a rich source of factors associated with marijuana use among youth.

PAGE 23

23 Figure 2 1. N MAPSS recruitment sites map F igure 2 2. Flow ch art for outcome measures for Chapters 3 and 4

PAGE 24

24 Figure 2 3. Flow chart for outcome measures for Chapter 5

PAGE 25

25 CHAPTER 3 INDIVIDUAL LEVEL FACTORS ASSOCIATED WITH CURRENT DAILY MARIJUANA USE AMONG US YOUTH WHO HAD INITIATED MARIJUANA USE 3.1 Background M arijuana u se is common among US youth. Adolescence, generally considered to range from ages 10 to 19 (Mackay and Duran, 2007), is a critical period for various types of drug initiation and experimentation including marijuana (SAMH SA, 2011). According to the 2016 Nat ional Survey on Drug Use and Health (NSDUH), about 6.5 % of youth aged 12 to 17 years reported marijuana use in the past month (SAMHSA, 2017 ). Another national survey among youth, the Monitoring the Future (MTF) survey of 2016 found the prevalence of past month marijuana use to be 5.4%, 14.0% and 22.5% among 8 th 10 th and 12 th grade students, respectively. These findings evidence an increase in marijuana use by grade level among adolescents. Moreover, nearly one half of high school students reported marijua na use at least once in their lifetime by the time they graduated from high school. Additionally, almost 70% of the high school seniors reported perceiving regular marijuana use as not harmful (Johnston et al., 2016). The use of marijuana among youth has generated significant public health concern due to adverse developmental, neurobiological, psychosocial and behavioral outcomes associated with its use. Research has consistently found these adverse outcomes among youth who endorse regular or heavy marijua na use (Azofeifa, 2016) Since brain development continues until early adulthood exposure to tetrahydrocannabinol (THC), a major psychoactive ingredient in marijuana, during adolescence may cause structural and functional changes resulting in developmental and neurobiological alterations (Dinieri and Hurd, 2012; Lisdahl et al., 2014) Early a ge

PAGE 26

26 of marijuana initiatio n and regular marijuana use have also been associated with developing cannabis use disorder (Hall and Degenhardt, 2009) Further, o ther identified effects of long term or regular use range from poor educational outcomes, cognitive impairment, respiratory problems (e.g. chronic bronchitis), diminished satisfaction in life achievement to risk of mental problems such as schizophrenia, particularly among those with a p redisposition to such disorders (Volkow et al., 2014; Hall and Dege ngardt, 2014) Based on prior research, some individual level factors have been shown to predict marijuana use among youth. Sociodemographic factors such as age, sex, race/ethnicity, family characteristics and academic performance have been found to pred ict youth marijuana use. The MTF study clearly indicated that older adolescents are more likely to use marijuana compared to younger ones: rates of use increased by grade level among 8 th 10 th and 12 th grade students (Johnston et al., 2016). Regarding sex, males have been consistently found to have higher rates of marijuana use compared to females. However, sex differences in use substantially declined in the period from 1999 to 2013 (Johns on et al., 2015) Similar changes are observed in the association between race/ethnicity and marijuana use among adolescents. Until the early 1990s, white high school students were more likely to use marijuana than Blacks, but this trend narrowed substant ially by the early 2000s. Then by 2013, current marijuana use prevalence was higher among Blacks, Hispanics and American Indians compared to Whites (Johnson et al., 2015) Asian American s continued to have the lowest prevalence of marijuana use among population groups (Johnson et al., 2015; Shih et al., 2010).

PAGE 27

27 Other factors reduce or increase the risk of marijuana use among youth. For example, living with both parents is a protective fac tor (Mandara et al., 2011; Musick et al., 2008; Vogel et al., 2015) whereas having divorced/ separated parents (Mandara et al., 2011) and being raised without a biological father (Swift et al., 2008) are risk factors for youth marijuana use. Further, youth who used marijuana were more likely to report poor academic performance including poor school grades (Homel et al., 2014; Tu et al., 2008) and not completing high school (Horwood et al., 2010; Silins et al., 2014) co mpared to those who did not. Individual behavioral risk factors such as the use of tobacco, alcohol and other illicit drugs and age of marijuana initiation have also been shown to affect youth marijuana use. Youth who used marijuana compared to their count erparts were more likely to endorse use of tobacco (Guxens et al., 2007; Korhonen et al., 2008; Tu et al., 2008; Nel son et al., 2015) alcohol (Guxens et al., 2007; Korhonen et al., 2008; Perkonigg et al., 2008; Tu et al., 2008) and other drugs (Hall and Degenhardt, 2007). Early initiation of marijuana has also been associated with increased risk of regular marijuana use (Terry al., 2015) Moreover, marijuana using youth had higher likelihood of reporting symptoms of anxiety (Mar morstein et al., 2010) depression (Pa cek et al., 2012; Tu et al., 2008; Wright et al., 2016) conduct disorder (Wymbs et al., 2012) aggressive behavior (Farhat et al., 2011; Korhonen et al., 2008) and attention problems (King et al., 2011; Sibley et al., 2014) compared to their counterparts. While the majority of research on marijuana use among youth has been fo cused on understand ing the risk factors associated with lifetime or past 30 day use versus no

PAGE 28

28 use, to our knowledge no national studies specify individual level risk factors associated with current daily marijuana use and non daily use among those who have already initiated use. Research indicates that about 10% of those who initiate marijuana use become daily users (Hall and Degenhardt, 2009) As noted, most health consequences occur due to regular marijuan a use during adolescence and several epidemiological studies define regular marijuana use as daily or almost daily use (Hall and Degenhardt, 2009), it is important to focus on determining factors specific to current daily marijuana use among youth who have initiated marijuana use. Prior research shows that definitions of daily marijuana use vary between studies. The annual cross sectional national study on substance use, NSDUH, defines past month daily or almost daily marijuana use as use on 20 or more day s in the past 30 days while the MTF study defines daily use as use on 20 or more occasions in the past 30 days. Other studies have defined daily marijuana use as having used every day (Silins et al., 2014) or having used at least 6 days in a week (Conroy et al., 2016) In addition to daily use, studies have also chosen to measure heavy use, regular use or frequent use; some of these definitions are comparable with daily use. For instance, in one study, frequent use was defined as having used 20 to 39 tim es in the past 30 days (Johnson et al., 2015) ; however, other definitions of frequent use include having used on at least 3 days per week for at least 12 months (van der Pol et al., 2013) and having used at least 3 times in the past 30 days (Farhat et al., 2 011) Heavy use has been defined as having used 10 or more times in the past 30 days (Tu et al., 2008), having used on a monthly basis or more frequently (Griffin et al., 2002) and having used on at least 4 occasions in the past four weeks (Creemers et al ., 2010). Given this landscape,

PAGE 29

29 it is difficult to compare and replicate findings from different studies. Thus, risk factors specific to a standard definition of current daily marijuana use among youth remain unclear. Therefore, the main objective of this analysis was to determine the individual level factors that predict, among youth who had initiated marijuana use, current daily marijuana use and non daily use versus no current use (use on zero days in the past 30 days) Consistent with the NSDUH survey, we defined current daily marijuana use as having used the substance on 20 or more days in the past 30 days. We tested the same individual level factors that have been found to be associated with youth marijuana use in the literature to determine factors t hat predict current daily marijuana use. We hypothesized that individual level factors such as early age of initiation, poor grades, current use of tobacco, current use of alcohol, lifetime use of other drugs, having depressive and an xiety symptoms, conduc t disorder symptoms, and ADD/ADHD predict daily marijuana use among youth who had initiated marijuana use. Further, we hypothesized that alcohol and drug use in general would be related, such that lifetime and current use would be most predictive of daily marijuana use. 3.2 Methods Data were drawn from the National Monitoring of Adolescent Prescription Stimulants Study (N MAPSS; PI LB Cottler); a cross sectional study that aimed to determine the prevalence of and risk factors for use, misuse and diversion o f prescription stimulants among youth. N MAPSS recruited a total of 11,048 youth 10 to 18 years of age from shopping malls, movie theatres, arcades, libraries, skate parks, parks, sports and recreational centers using an entertainment intercept venue metho d from urban, suburban and rural areas of 10 US cities. Trained interviewers obtained

PAGE 30

30 verbal consent from each participant for anonymous interviews and the survey was s elf administered among 13 to 18 year olds. Participants took about 20 30 minutes to comp lete the survey. The survey was conducted in four waves from 2008 to 2011. Participants received a remuneration of a gift card worth $10 from a national electronics store upon survey completion. The study protocol was approved by Washington University in S t. Louis and the University of Florida Institutional Review Boards (Cottler et al., 2013; Lasopa et al., 2015; Striley et al., 2017; Wang et al., 2015) 3.2.1 Samp le The sample was restricted to 13 to 18 year olds (n=9,642) due to low rates of self reporte d marijuana use among 10 to 12 year olds. Participants were asked if they urrent anal yses (n=3,155; 32.72% ). 3.2.2 Measures 3.2.2.1 Current marijuana u se The main outcome of interest in this study was current marijuana use (past 30 day use) among youth who had previously initiated marijuana use. Participants who reported having ever used marijuana were asked about their age they first used marijuana. They were further asked two days, three to five days, six t o nine days, 10 to 19 days, 20 to 29 days and all 30 to nine days, and 10 to 19 da

PAGE 31

31 3.2.2.2 Sociodemographic f actors Sociodemographic variables included in the analyses were self reported age, sex, race/ethnicity, self reported school grades, and family structure. Age remained as a continuous variable; sex was categorized into male or female; race/ethnicity was categorized into Non Hispanic White, Non Hispanic Black, Hispanic, Asian/Asian American and other races (Alaskan Native/Eskimo, American Indian, Middle Eastern, Pacific Islander, Biracial or multiracial); self reported grades in school variable was egorized into lived with both mother and father, lived with either mother or father, or lived with others (foster parents, relatives or other). 3.2.2.3 Behavioral f actors Behavioral variables included in the analyses were current use of tobacco, current us e of alcohol, lifetime use of other drugs, and age of marijuana initiation. To assess the current use of tobacco, participants were first asked if they ever smoked a day or some days. Youth who responded that they currently smoked cigarettes every day or some days were categorized as current tobacco users. Similarly, to assess current alcohol use, participants were asked if they had ever had a beer, a glass of wine or any how many days they drank in the last 30 days. Those who responded one or more days were categorized as current alcohol users. Lifetime use of other drugs was assessed by asking if they had ever tried cocaine or crack, heroin, club drugs like ecstasy,

PAGE 32

32 hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants. Youth who reported having ever used any of these drugs were categorized as having lifetime other illegal drug use. 3.2.2.4 Mental health f actors Mental health problems included in the analyses were having ADD/ ADHD, and symptoms of anxiety, conduct disorder and depression. Having been diagnosed with or ever told you or your parents that you have Attention Deficit Disorder (ADD) or Attention Deficit Hyperactivity worried or stressed for 6 months or more. Similarly, conduc t disorder symptoms were assessed using two items: if youth had ever (1) gotten into a lot of trouble at home or at school or run away from home overnight or (2) used or threatened someone with a weapon. Depression symptoms were assessed using two items: i n the last 12 months, if youth had two weeks or more when they (1) lost interest in things or (2) felt down or depressed. Overall, 55 youth had missing data for variables of interest; therefore, we deleted those observations leaving complete data for 3 100 youth 3.3 Statistical Analyses Descriptive statistics of sociodemographic, behavioral and mental health characteristics were calculated by current daily use, current non daily use and no current use using F tests and chi square tests with the significanc e level set at a p value of < 0.05. Then, unadjusted odds ratios were estimated by creating 13 bivariate logistic regression models for each individual level factor and the outcome of interest. We then assessed for multicollinearity to identify redundant v ariables for omission using a

PAGE 33

33 regression model that included all variables to estimate Variance Inflation Factors (VIF). A VIF coefficient below 10 was desired (Davis et al., 1986). The variables included in our model indicated VIF coefficients less than t wo, resulting in low to no multicollinearity. Further, a multinomial logistic regression model was conducted to determine the effects of individual level factors associated with current daily marijuana use and current non daily marijuana use compared with no current marijuana use (reference group). All the individual level factors that were pre selected are important variables in marijuana use among youth in the literature and thus, were included in the full regression model Adjusted odds ratios with 95% c onfidence intervals were reported. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). 3.4 Results Table 3 1 shows the sociodemographic, behavioral and mental health characteristics by current marijuana use among US youth 13 to 18 years of age who had initiated marijuana use. Out of 3,100 youth who reported lifetime marijuana use, 41.00% reported no current use, 44.71% reported current non daily use and 14.29% reported current daily use. Among youth who had initiated marij uana use in N MAPSS, the reported mean age was 16.33 years; 53.55% were male, 44.19% were NH White, 44.13% lived with both mother and father, and 38.23% reported their grades in school as Cs or worse. Moreover, 29.35% reported current use of tobacco, 60.35 % reported current use of alcohol and 37.68% reported lifetime use of other drugs. The reported mean age of marijuana initiation was 13.99 years. Further, 20.55% reported having ADD/ADHD and 35.39%, 23.23%, and 32.32% reported having symptoms of anxiety, c onduct disorder, and depression, respectively.

PAGE 34

34 Bivariate analyses showed that sociod emographic factors such as sex, family structure, and self reported grades in school were found to be significantly associated w ith current marijuana use Compared to non c urrent marijuana users (48.54%), current daily (72.23%) and current non daily marijuana use rs (52.16%) were more likely to be mal es (p<.05). Compared to non current marijuana users (32.65%), current daily (53.72%) and current non daily marijuana users (38. 38%) were also more likely to report Cs or worse grades in scho ol ( p<.05). Further, non current (12.98%) and current non daily marijuana users (12.34%), current daily users (22.57%) were more likely to report living with foster parents, relatives or others ( p<.05). Age and race were not significantly associated w ith current marijuana use (Table 3 1). All behavioral factors, including current tobacco and alcohol use, lifetime other drug use, and age of marijuana initiation, were significantly associated w ith current marijuana use Compared to non current marijuana users (19.28%), current daily (57.79%) and current non daily users (29.51%) were more likely to report current tobacco u se (p<.05). Similarly, compared to non current marijuana users (44.77%), curre nt daily (69.41%) and current non daily users (76.75%) were more likely to report current alcohol use (p<.05) and compared to non current marijuana users (25.81%), current daily (36.58%) and current non daily users (75.17%) were more likely to report lifet ime use of other illicit drugs (p<.05). Further, current daily marijuana use rs (12.84 years) compared to current non daily use rs (14.09 years) and no n current use rs (14.29 years) were more likely to report early age of marijuana initiation (p<.05). Among t he mental health factors, both having been diagnosed with ADD/ADHD and with symptoms of conduct disorder were significantly associated w ith current marijuana use Compared

PAGE 35

35 to non current (17.55%) and current non daily marijuana users (18.98%), current dail y marijuana use rs (34.09%) were more likely to report having been diagnosed with ADD/A DHD ( p<.05) and compared to non current (19.75%) and current non daily marijuana users (20.63%), current daily marijuana use rs (41.31%) were more likely to report symptom s of conduct disor der ( p<.05). However, symptoms of anxiety and depression were not significantly associated with current ma rijuana use (Table 3 1). Results from the multinomial logistic regression indicated that age, sex, race, family structure and self reported school were significant predictors of current marijuana use With each year increase in age, youth were 24 .00 % more likely to be current daily marijuana users compared to no n current users, holding all other variables constant. In contrast, with e ach year increase in age, youth were 13 .00 % less likely to be current non daily marijuana users compared to no n current users, holding all other variables constant. Compared to no n current marijuana users, current daily and current non daily use rs were mor e likely to be male ( AORs= 2.64 and 1.29), NH Black compared to NH White (AORs= 1.63 and 1.64), and report living with either mother or father compared to living with both mother and father (AORs= 1.53 and 1.30) and Cs or worse grades in school (AORs= 1.49 and 1.23). Moreover, compared to no n current marijuana users, current daily marijuana users were 1.51 times more likely to be of other rac es versus NH White and 1.54 times more likely to report living with foster parents, relatives or others versus living with both mother and father (Table 3 2). Both c urrent daily and current non daily marijuana users were 2.55 and 1.41 times more likely to endorse curren t tobacco use and 2.78 and 3.04 times more likely to endorse al cohol use, respectively compared to no n current marijuana users. Moreover,

PAGE 36

36 current daily and current non daily marijuana users were 3.99 and 1.34 times more likely to endorse lifetime other drug use, respectively compared to no n current marijuana users. In addition, with each year increase in age of marijuana initiation, youth were 23 .00 % less likely to be current daily marijuana users compared to no n current users, holding all other variables constant. Having been diagnosed with ADD/ADHD was also a significant predictor of current daily mariju ana use; symptoms of anxiety, conduct disorder and depression did not significantly predict current marijuana use among youth who had initiated marijuana use (Table 3 2). 3.5 Discussion To our knowledge, this wa s one of the first studies to examine indivi dual level factors associated with current daily marijuana use among US youth who had initiated marijuana use, using a national sample. We used a specific current daily marijuana use definition that is consistent with the national survey NSDUH allow ing c omparison across studies in the future. In this study, almost 30% of the youth reported marijuana use in their lifetime; nearly 15% reported current daily marijuana use, whereas a majority of youth reported either current non daily marijuana use or no curr ent use. In a review article, Hall and Degenhardt (2009) estimated that approximately 10% of those who initiate marijuana become daily users (Hall and Degenhardt, 2009) The prevalence of current daily marijuana use among our sample of youth who initiated marijuana was higher than the prevalence specified in this review article. This difference in prevalence may be because the review focused on individuals of all age groups while our study only focused on 13 to 18 year olds. Prior research on marijuana use among youth demonstrated that older adolescents had higher likelihoods of using marijuana compared to younger

PAGE 37

37 adolescents (Johnston et al., 20 16). The results of this study we re consiste nt in that among youth who had initiated marijuana use, current daily marijuana use was associated with increasing age. However, current non daily marijuana use was inversely associated with increasing age. These findings suggest that with increasing age, youth are more likely to be at the extreme ends; either they are more likely to use marijuana daily or not reinitiate use at all once they stop. Research also indicated that compared to females, males had higher odds of marijuana use (Johnson et al., 2015 ; Johnston et al., 2016 ) and cannabis use disorder (Haberstick et al., 2014; Hasin et al., 2015) in all age groups including adolescents. Consistent with other stu dies, we found that among youth who initiated marijuana use, current daily and current non daily marijuana users were more likely than no n current users to be male. Similar to findings from the stud y by Johnson and colleagues (2015) current daily and curr ent non daily marijuana users were more likely to be NH Black compared to no n current users. However, unlike their findings on Hispanics, we fo und no significant difference in current marijuana use among Hispanics and Asians. However, current daily marijua na users were more likely to be of other races compared to no n current marijuana users. Additionally, studies have found living with both parents to be protective of youth marijuana use (Mandara et al., 2011; Musick et al., 2008; Swift et al., 2008; Vogel et al., 2015) Our findings were similar for youth who had initiated marijuana use; both current daily and non daily marijuana users were more likely to liv e with a single parent; current daily users were also more likely to live with foster parents, relatives or others. Several investigations have demonstrated a positive relationship between poor academic performance and youth marijuana use (Homel et al., 2014; Tu et al., 2008) This

PAGE 38

38 association was consistent with both current daily and non daily marijuana users reporting lower grades in school than non current users. In terms of the relationsh ip between marijuana use and the use of other substances, we found that among youth who had initiated marijuana use, more than a quarter reported current tobacco use and lifetime use of other drugs and mo re than half reported curr ent alcohol use. Current use of tobacco and alcohol and lifetime use of other drugs have been found to be the most important factors influencing the risk of current daily and current non daily marijuana use in this study. This finding was consistent with the body of research exhibiting a positive association between the use of alcohol, tobacco, and other drugs and marijuana use (Guxens et al., 2007; Hall and Perkonigg et al., 2008; Tu et al., 2008) and supoorted our hypothesis I t is important to note that the like lihood of current daily marijuana use was lower than no current use with increasing age of marijuana initiation. Sagar and colleagues (2015) had also found higher leve ls of marijuana use among early onset mar ijuana smokers compared to late onset smokers, s upporting our findings (Sagar et al., 2015). Considering self reported mental health conditions, current daily marijuana users were more likely to report having been diagnosed with ADD/ADHD compared to no n current marijuana users T his association was mar ginally significant. Although a significant proportion of youth in the study sample reported having symptoms of anxiety, conduct disorder and depression, in contrast to our hypothesis, this study did not find any association between these mental health ch aracteristics and current marijuana use

PAGE 39

39 This may have had to do with the nature of our questions for these mental health problems. They did not provide the detail necessary to diagnose these disorders These study findings must be interpreted in light of some limitations. First, N MAPSS was a cross sectional study that limited our ability to infer temporal associations. Second, the study did not capture patterns of use among marijuana users prior to the past 30 day s. Third, the outcome variable, current m arijuana use was divided into three groups, derived from a single question with seven responses and this might have misclassified some youth. Since our focus was current daily marijuana use, we mimicked the same definition used by the NSDUH study. Categor izing use of marijuana on one or two days and 10 to 19 days in the past 30 days in the same category might have resulted in misclassification and caused bias in the study However, we did not find significant empirical differences in between those groups. We wanted to ensure that we had sufficient sample size in each group and three groups also captured the inherent variation in the data with the least loss of information. Fourth, N MAPSS did not assess other important individual level factors including dos age a nd frequency of marijuana use and clinica l diagnosis of cannabis use disorder and mental health disorders While there were some weaknesses, the study covered a wide range of important variables including sociodemographic, mental, and behavioral fac tors that have been associated with youth marijuana use. N MAPSS is one of the few national studies that has a large and diverse sample of youth with a significant proportion of lifetime marijuana users (much higher than school or household based studies) (SAMHSA, 2015; Johnston et al., 2016) and it provided an especially rich source for

PAGE 40

40 understanding the risk and protective factors associated with marijuana use. In addition, the use of entertainment venue intercept method of sampling enabled recruitment of youth not enr olled in and absent from school and home schooled youth. Such youth are mostly missing in school based national studies. Moreover, recruiting youth from social venues also might have reduced social desirability bias since parental approval wa s not required for participation in the study and the survey was anonymous 3.6 Conclusion Prevention efforts should be focused on all three marijuana using groups no current use, current non daily use and current daily use. Among prior initiators, you th with no current marijuana use should receive interventions to reduce the likelihoo d of reinitiating use Youth with current non daily marijuana use should be encouraged to reduce the frequency of use or to stop use by increasing the salienc e of conseque nces of daily use. Based on our findings, most of the individual level factors that predicted current daily and current non daily marijuana use did not differ (being male, NH Black, living with single parent, reporting poor academic performance, current us e of tobacco and alcohol a nd lifetime use of other drugs); however, the odds were higher among current daily users. Moreover, t here are some additional risk factors specific to current daily marijuana use. Such factors we re older age, being of other races including Alaskan Native/Eskimo, American Indian, Middle Eastern, Pacific Is lander, Biracial or multiracial living with foster parents, relatives and others, early age of marijuana initiation and having been diagnosed with ADD/ADHD Prevention efforts sho uld target youth with these risk factors in order to reduce daily marijuana use associated with adverse consequences including the risk of cannabis use disorder, injuries, respiratory problems, psychotic disorders, and changes in brain structure and functi oning Future

PAGE 41

41 studies should investigate risk factors associated with daily marijuana use using longitudinal studies which can help determine risks for progression from marijuana initiation to daily marijuana use among youth.

PAGE 42

42 Table 3 1. Socio demographic, behavioral and mental health characteristics by current marijuana us e among youth 13 to 18 years of age who had initiated marijuana use in N MAPSS (n=3,100 ) Characteristic Total (n=3,100) No current use (n=1271; 41.00%) Current non daily use (n=1386; 44. 71%) Current daily use (n=443, 14.29%) p value Socio demographic Age (Mean SD) 16.33 1.36 16.33 1.37 16.21 1.37 16.71 1.28 0.1357 Male gender 1660 (53.55%) 617 (48.54%) 723 (52.16%) 320 (72.23%) <.0001 Race/Ethnicity NH White NH Black Hispanic Asian/Asian American Others 1370 (44.19%) 505 (16.29%) 739 (23.84%) 122 (3.94%) 364 (11.74%) 565 (44.45 %) 194 (15.2 6%) 316 (24.86 %) 60 (4.72%) 136 (10.70%) 6 0 6 (43.7 2%) 245 (17.68 %) 326 (23.52 %) 45 (3.25%) 164 (11.83%) 199 (44.9 2%) 66 (14.90 %) 97 (21.90 %) 17 (3.84%) 64 (14.45%) .1484 Family Structure <.0001 Lived with both mother and father 1368 (44.13%) 622 (48.94%) 599 (43.22%) 147 (33.18%) Lived with either mother or father 1296 (41.81%) 484 (38.08%) 616 (44.44%) 196 (44.24%) Lived with others 436 (14.06%) 165 (12.98%) 171 (12.34%) 100 (22.57%) Self reported school grades (Cs or worse) 1185 (38.23%) 415 (32.65%) 532 (38.38%) 238 (53.72%) <.0001 Behavioral Current tobacco use 910 (29 .35%) 245 (19.28%) 409 (29.51%) 256 (57.79%) <.0001 Current alcohol use 1871 (60.35%) 569 (44.77%) 962 (69.41%) 340 (76.75%) <.0001 Lifetime other drug use 1168 (37.68%) 328 (25.81%) 507 (36.58%) 333 (75.17%) <.0001 Age of marijuana initiation (Me an SD) 13.99 1.96 14.29 1.81 14.09 1.86 12.84 2.25 <.0001 Mental Health ADD/ADHD 637 (20.55%) 223 (17.55%) 263 (18.98%) 151 (34.09%) <.0001 Anxiety symptoms 1097 (35.39%) 449 (35.33%) 478 (34.49%) 170 (38.37%) .3293 Conduct disorde r symptoms 720 (23.23%) 251 (19.75%) 286 (20.63%) 183 (41.31%) <.0001 Depressive symptoms 1002 (32.32%) 419 (32.97%) 429 (30.95%) 154 (34.76%) .2676 Race/Ethnicity Others: Alaskan Native/Eskimo, American Indian, Middle Eastern, Pacific Islander, and bi racial or multiracial; Lived with others: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifetime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inh alants

PAGE 43

43 Table 3 2. Multinomial logistic regression model of socio demographi c, behavioral and mental health characteristics for current marijuana use among youth 13 to 18 years of age who had initiated marijuana use in N MAPSS (n=3,100 ) Characteristic C urrent non daily use versus No current use AOR (95% CI) Current daily use versus No current use AOR (95% CI) Socio demographic Age 0.87 (0.81, 0.93) 1.24 (1.11, 1.37) Sex (Male versus female) 1.29 (1.10, 1.52) 2.64 (2.02, 3.46) Race NH Whit e Ref Ref NH Black 1.64 (1.28, 2.10) 1.63 (1.10, 2.42) Hispanic 1.02 (0.83, 1.25) 0.90 (0.65, 1.25) Asian/ Asian American 0.78 (0.51, 1.19) 1.15 (0.60, 2.21) Others 1.24 (0.94, 1.62) 1.51 (1.02, 2.23) Family Structure Living with both mother and father Ref Ref Living with either mother or father 1.30 (1.09, 1.55) 1.53 (1.15, 2.03) Living with others 1.05 (0.81, 1.36) 1.54 (1.07, 2.22) Self reported school grades (Cs or worse) 1.23 (1.03, 1.47) 1.49 (1.15, 1.93) Behavi oral Current tobacco use 1.41 (1.15, 1.74 ) 2.55 (1.92, 3.38) Current alcohol use 3.04 (2.55, 3.61) 2.78 (2.09, 3.71) Lifetime other drug use 1.34 (1.11, 1.62) 3.99 (3.00, 5.31) Age of marijuana initiation 1.01 (0.95, 1.06) 0.77 (0.72, 0.83) Mental Health ADD/ADHD 1.06 (0.85, 1.31) 1.35 (1.01, 1.81) Anxiety symptoms 0.96 (0.80, 1.15) 0.80 (0.61, 1.06) Conduct disorder symptoms 0.81 (0.66, 1.00) 1.16 (0.87, 1.54) Depressive symptoms 0.83 (0.69, 1.00) 0.89 (0.67, 1.19) Race/Ethn icity Others: Alaskan Native/Eskimo, American Indian, Middle Eastern, Pacific Islander, and biracial or multiracial; Lived wi th others: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifetime use of cocaine or crack, heroi n, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants Bold: Significant association

PAGE 44

44 CHAPTER 4 SEX AND RACE/ETHNICITY SPECIFIC INDIVIDUAL LEVEL FACTORS ASSOCIATED WITH CURRENT DAILY MARIJUANA USE AMONG YOUTH 4.1 Background Marijuana use among youth poses a significant public health concern, with evidence indicating adverse developmental, neurobiological, psychosocial and behavioral consequences associated with its use (Volkow et al., 2014; Hall and Degenhar dt, 2014). Recent data has shown that despite the changing of state marijuana laws in the US, the prevalence of marijuana use in the past year among 8 th and 10 th grade students has decreased while that among 12 th grade students had not changed in 2016 com pared to 2011 (Johnston et al., 2016). Although the prevalence of use did not seem to increase among high school students in 2016, the proportion of youth who perceived regular marijuana use as risky continued to decrease among all grades and almost 40% of the high school students had tried marijuana by the end of high school (Johnston et al., 2016). Therefore, the potential for increase in marijuana use among adolescents due to decreased perception of risk from regular marijuana use still exists. Researc h suggests a number of in dividual level risk factors ar e associated with youth marijuana use Such factors include age (Johnston et al., 2016), sex (Johnson et al., 2015), race/ethnicity (Johnson et al., 2015), family composition (Mandara et al., 2011; Mus ick et al., 2008; Vogel et al., 2015; Swift et al., 2008), academic performance (Homel et al., 2014; Tu et a l., 2008; Horwood et al., 2010; Silins et al., 2014) use of tobacco (Guxens et al., 2007; Korhonen et al., 2008; T u et al., 2008; Nelson, Ryzin and Dishion, 2015) use of alcohol (Guxens et al., 2007; Korhonen al., 2011; Perkonigg et al., 2008; Tu et al., 2008) and use of other drugs (Hall and

PAGE 45

45 Degenhardt, 2007; Moss et al., 2014). Moreove r, early age of marijuana initiation has been associated with elevated levels of marijuana use (Griffin et al., 2010; Terry Further, marijuana use in youth has been associated with mental health problems such as symptoms of ADHD (King et al., 2011; Sibley et al., 2014; Kousha et al., 2012; Barkley et al., 2004) anxiety (Marmorstein et al., 2010) conduct disorder (Wymbs et al., 2012) and depression (Pacek et al., 2012; Tu et al., 2008; Wright et al., 2016) While a majority of youth who initiate marijuana u se smoke marijuana occasionally, a subset of youth use it more fre quently (Adlaf and Paglia, 2003; Tu et al., 2008). According to the 2016 Monitoring the Future (MTF) survey, about 0.7% of the 8 th graders, 2.5% of the 10 th graders and 6.0% of the 12 th graders reported daily marijuana use (de fined as use on 20 or more occasions in the past 30 days). Further, the proportion of daily marijuana users is higher than that of daily cigarette users (Johnston et al., 2016). Characteristics of daily marijuana users identified from MTF were being male and white, non college bound seniors, from broken homes and having poor school grades (Johnston, 1981). Findings from our recent research have indicated that compared to youth who endorsed no current marijuana use those who endorsed current daily ma rijuana use had a higher likelihood of being older, mal e, Non Hispanic (NH) Black and individuals of other races compared to NH White and having lived without both parents (Chapter 3). Moreover, youth who endorsed current daily marijuana use were also more likely to report poor academic performance, current us e of tobacco and alcohol, lifetime use of other drugs, early age of marijuana initiation and

PAGE 46

46 having been diagnosed with ADD/ADHD Further research on risk factors for marijuana use in youth is still needed A key area of interest for research on marijuana use in youth is sex differences in risk factors for use. Prior research has found that patterns of substance use, frequency of use and use disorder are not equally distributed by sex (McCabe et al., 2007). Males are more likely to use all kinds of illicit drugs compared to females. This disproportion might explain the higher number of emergency department visits and overdose deaths among males compared to females (SAMHSA, 2014; NIDA, 2016). Similarly, like other illicit drug use among youth marijuana use a l so exhibits differences by sex. The National Youth Risk Behavior Survey (YRBS) demonstrated that males we re significantly more likely to report higher rates of marijuana use than females. How ever, this difference markedly declined from 1999 to 2013 (Johns on et al., 2015). In 2015, males had a higher prevalence of lifetime marijuana use than females among Black adolescents only; however, the prevalence of current marijuana use was higher in males than in females for all adolescents. Moreover, males were alm ost twice as likely as females to report daily marijuana use (Johnston, 1981) and 1.4 to 1.8 times more likely to report cannabis abuse or dependence compared to females (Haberstick et al., 2014). Further females who reported marijuana use 10+ times in th e past 30 days were more likely to have poor mental health and academic performance (Tu et al., 2008). Other correlated factors do not differ by sex. For example, a cross sex, current marijuana users compared to non users were more likely to report use of other substances, conduct problems such as fighting or carrying weapons, depressed mood and being less likely to participate in extracurricular activities (Schepis et al., 2011). In

PAGE 47

47 another study, both male s and female s who use d marijuana 10+ times in the past 30 days were more likely to be older and to report poor economic status and frequent use of tobacco and alcohol. Given the available evidence for both differences and similarities in marijuana use risk factors by sex it is important to determine if s ex differences in other individual level risk factors for daily marijuana use exist in youth. If so, prevention efforts need to account for such differences. Similarly, racial differences i n marijuana use in youth have been identified (Keyes et al., 2015; Keyes et al., 2017; Johnson et al., 2015; Wu et al., 2015). Throughout the 1900s in the US, White high school students had a higher likelihood of using marijuana than Blacks and Hispanics (Johnson et al., 2015) This trend has changed over time. Presently, Black students (27.1%) are at higher risk for marijuana use, followed by Hispanic (24.5%) and White students (19.9%) as reported in the 2015 YRBS (Kann et al., 2016). In addition, Asian youth tend to have the lowest prevalence of marijuana use among all r aces (Shih et al., 2010). Similar trends have been reported by the Monitoring the Future (MTF) study (Miech et al., 2016) with some findings indicating an early age of marijuana initiation among minority youth (Kosterman et al., 2000). Racial/ethnic varia tion in individual beliefs, family or peer factors, and genetic polymorphisms may contribute to adolescent substance use of all types (Shih et al., 2010). Individual level factors as they vary by race/ethnicity are understudied; they may modify the effect of race on substance use including daily marijuana use. However, a small group of studies has begun to indicate their importance. An increased risk of substance use among Hispanic youth has been associated with individual level factors including low resist ance self

PAGE 48

48 their belief toward and perception of the consequences of substance use; family and school level factors did not play a role. Keyes and colleagues (2017) used two individual level fact ors (sex and parental education) to predict marijuana use among 8 th 10 th and 12 th grade students by race/ethnicity and found higher parental education was associated with increased current marijuana use among Blacks compared to Whites. While the patterns of marijuana use vary by race/ethnicity, other risk factors may continue to contribute to such patterns of use. Being from a minority group may increase the likelihood of experiencing drug related problems. For instance, Black and Hispanic you th are at hig her risk of being arrested or facing severe penalties for marijuana related offenses even though marijuana use rates are comparable to those of Whites (American Civil Liberties Union, 2013; Wu et al., 2013). A study among African American middle and high s chool students found that marijuana users were more likely to be male and engage in risky behaviors including getting in trouble at school and with police and to attend parties where alcohol and other drugs were available compared to non users (Vidourek et al., 2017). Moreover, Black and Hispanic youth were at an increased risk of experiencing alcohol related problems compared to White youth (Witbrodt et al. 2014). Further, Ellickson and Morton (1999) reported that among marijuana users, Hispanics had a hig her likelihood of using cocaine and heroin compared to Whites or Blacks (Ellickson and Morton, 1999). Given this evidence suggesting their importance, there is a need to examine other individual level factors associated with daily marijuana use by race/eth nicity. Findings from such studies may help target particular intervention elements for the minority population.

PAGE 49

49 With a majority of studies being conducted among marijuana users versus non users, it is also critical to study the variations among daily user s versus non daily users by sex and race/ethnicity W e aimed to determine if different sets of predictive variables were of utility in predicting daily marij uana use by sex and race/ethnicity Therefore, the current analyses aimed to examine a wide variety of individual level factors that predict current daily marijuana use among youth who had initiated marijuana use by (1) sex and (2) race/ethnicity separately. We hypothesized that across sex, increasing age, current use of tobacco and alcohol, lifetime us e of other drugs, early age of initiation and symptoms of conduct disorder and depression predict daily marijuana use Moreover, based on the literature, we hypothesized that the predictive variables would differ by sex, with poor grades and mental health problems being more predictive among females than males. We also hypothesized that among African American youth, males would have a higher likelihood of current daily marijuana use than females, and that conduct disorder symptoms would be more predictive of current daily marijuana use than among other races. 4.2 Methods The study sample was obtained from the National Monitoring of Adolescent Prescription Stimulants Study (N MAPSS), a cross sectional study that was conducted across 10 US cities. N MAPSS aimed to examine the prevalence of and risk factors associated with use, misuse and diversion of prescription stimulants among youth. Trained interviewers recruited 11,048 youth between the ages of 10 and 18 years using an entertainment intercept venue met hod and participants completed the 20 to 30 minute long survey. The assessment included questions about sociodemographic, behavioral and mental health characteristics, and use of alcohol, tobacco, prescription

PAGE 50

50 drugs, marijuana and other illicit drugs. Data was collected in four waves from fall 2008 to spring 2011. The detailed methodology of N MAPSS has been reported elsewhere (Cottler et al., 2013; Lasopa et al., 2015; Striley et al., 2017; Wang et al., 2015). The study was approved by the Washington Unive rsity Institutional Review Board in St. Louis and the University of Florida Institutional Review Board. 4.2.1 Sample Due to low rate of self reported marijuana use, youth between 10 and 12 years were excluded from the analyses. Out of 9,642 13 to 18 year o lds analyses only included you th who reported having ever used marijuana (n=3,155; 32.72% ). Sex was dichotomized into male and female ; 1,692 males and 1,463 females we re included in these analyses. Categories for r ace/ethnicity were Non Hispanic (NH) Whit e, NH Black, and Hispanic; 1,382 NH Whites, 519 NH Blacks, and 745 Hispanics youth were included in these analyses. We did not study individual level factors specific to Asian/Asian American and other race/ethnicity such as Alaskan Native/Eskimo, American Indian, Pacific Islander, Middle Eastern and biracial and multiracial because the sample si zes were small for each and grouping them together would not be appropriate due to cultural differences. However, N MAPSS participants reporting these races/ethnici ties are included in the sex specific models. For the sex specific models, all race included models, 32 males and 63 females had missing data for variables of interest, therefore, we deleted those observations leaving complete data for 1,660 males and 1,40 0 females. Rather than imputing, we chose to delete these incomplete cases. For the race specific models, 12 NH Whites, 14 NH Blacks and 6 Hispanics had missing data for variables of interest; therefore, we deleted those observations leaving complete data for 1,370 NH Whites, 505 NH Blacks and 739 Hispanics.

PAGE 51

51 4.2.2 Measures 4.2.2.1 Current marijuana u se Current marijuana use was the main outcome variable of the study. Youth who reported having used marijuana in their lifetime were asked their first age of u se of three to five days, six to nine days, 10 to 19 days, 20 to 29 days and all 30 days. Since the focus of our study was current daily marijuana use, we categorized use on 20 or use on one to 19 days in the last and use of marijuana on zero days in the l ast 30 d ays as Current daily marijuana use was defined as use on 20 or more days in the past 30 days which is consistent with the National Survey on Drug Use and Health (NSDUH) definition. 4.2.2.2 Individual level f actors Potential individual le vel factors selected fell into three domains: sociodemographic, behavioral and mental health factors. Sociodemographic factors included age (13 to 18 years) family structure (lived with both parents, lived with either mom or dad, and lived with others inc luding foster parents, relatives or other ) and self reported grades in school (As or Bs and Cs or worse). Behavioral factors included current tobacco use, current alcohol use, and lifetime use of other drugs ( any yes to cocaine or crack, heroin, club drug s like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants versus none ). Current tobacco use was assessed using two items: (1) Have you ever smoked a cigarette? Participants who

PAGE 52

52 wed up by asking (2) Do you still smoke cigarettes every day or some days? Participants who reported smoking cigarettes every day or some days were classified as having current tobacco use. Similarly, current alcohol use was assessed using two items: (1) Have you ever had a beer, a glass of wine, or any other alcoholic drink, not just a sip? Participants who responded In the last 30 days, on how many days did you drink alcohol? Participants who rep orted drinking alcohol on one or more days in the past 30 days were categorized as having current alcohol use. Finally, mental health factors selected included having been diagnosed with ADD/ADHD and symptoms of anxiety, conduct disorder or depression. T o assess having ADD/ADHD, participants were asked if a doctor had ever told them or their parents that they had Attention Deficit Disorder (ADD) or Attention Deficit Hyperactivity Disorder (ADHD). Symptoms of anxiety were assessed by asking if they had eve r felt worried or stressed for six months or more. Symptoms of conduct disorder were examined using two items: (1) Have you ever gotten into a lot of trouble at home or at school or ran away from home overnight? (2) Have you ever used or threatened some one with a weapon? were categorized as having conduct disorder symptoms. Likewise, symptoms of depression were assessed using two items: (1) In the last 12 months, have you had two weeks or mo re when you lost interest in things? (2) In the last 12 months, have you had two weeks or more when you felt down or depressed? Participants who symptoms.

PAGE 53

53 4.3 Statistical An alyses D escriptive statistics were calculated for the three dependent variable categories ( current daily marijuana use, current non daily use and no current use ) stratified b y sex and race/ethnicity. C hi square tests or Fisher tests and F tests with the si gnificance level set at a p value of < 0.05 compared factors by sex and race/ethnicity Unadjusted odds ratios were calculated using bivariate logistic regression models for each individual level factor. Multicolli nearity was assessed to identify redundant variables using a regression model that included all variables by estimating Variance Inflation Factors (VIF) Since all variables included in the model showed a VIF of less than 2 indicating no multicollinearity, we included all the variables in the mode ls. Further, multinomial logistic regression models were used to examine sex and race/ethnicity specific individual level factors associated with current marijuana use status Final adjusted odds ratios with 95% confidence intervals were reported for each category of sex and race/ethnicity. For factors that were significant in one or the other model by sex or race/ethnicity we tested for interaction. SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for all analyses. 4.4 Results 4.4.1 Sex Diffe rences Among 1,660 males, 19.28% endorsed current daily marijuana use; 43.55% endorsed current non daily marijuana use and 37.17% endorsed no current marijuana use. Bivariate analyses showed that among males, current marijuana use status was significantly associated with race/ethnicity, family structure, grades in school, current tobacco use, current alcohol use, lifetime other drug use, age of marijuana initiation, having been diagnosed with ADD/ADHD, and symptoms of conduct disorder (Table 4

PAGE 54

54 1). Multinomi al logistic regression indicated among males, current daily marijuana users compared to no n current users were 1.21 times more likely to be older and 2.14 and 4.15 times more likely to report current tobacco use and lifetime other drug use respectively C urrent daily marijuana use was 23.00% less likely with increasing age of marijuana initiation compar ed to no current use C ompared to no current use c urrent non daily marijuana use was 10.00% less likely with increasing age and such users were also 1.68 t imes more likely to be of other races ( Alaskan Native/Eskimo, American Indian, Pacific Islander, Middle Easte rn and biracial and multiracial) and 28 .00 % and 25 .00 % less likely to report symptoms of conduct disorder and depression respectively Moreover, a mong males, both current daily and current non daily marijuana users compared to no n current users were more likely to be NH Black, and to report Cs or worse grades in school, and current alcohol use (Table 4 2). Among 1,400 females, 8.54% endorsed current daily marijuana use; 46.04% endorsed current non daily marijuana use and 45.42% endorsed no current marijuana use. Bivariate analyses showed that among females, current marijuana use status was significantly associated with the family structure, grades in school, current tobacco use, current alcohol use, lifetime other drug use, the age of marijuana initiation, having ADD/ADHD, and symptoms of conduct disorder and depression (Table 4 1). Multinomial logistic regression indicated among females, current dail y marijuana users compared to no n current users were 1.83 times more likely to report having ADD/ADHD. Current daily marijuana use was 23 .00 % less likely with increasing age of marijuana initiation compared to no current use Current non daily marijuana us e was 15 .00 % less likely with increasing age compared to no current use Moreover, among females, both

PAGE 55

55 current daily and current non daily marijuana users compared to no n current users were more likely to report living with a single parent, and to endorse current tobacco and alcohol use and lifetime other drug use (Table 4 2). These differences between males and females were then tested for significance. No significant interactions were reported between sex and grades in school and sex and family structure in p redicting current marijuana use. The only significantly different odds ratio between the sexes was in self reporte d ADD/ADHD diagnosis, with females with ADD/ADHD more at risk of curre nt daily marijuana use than males 4.4.2 Racial /E thnic Differences Among 1,370 NH Whites, 14.53% endorsed current daily marijuana use; 44.23% endorsed current non daily marijuana use and 41.24% endorsed no current marijuana use. Bivariate analyses showed that among NH Whites, current marijuana use status was significantl y associated with sex, family structure, grades in school, current tobacco use, current alcohol use, lifetime other drug use, the age of marijuana initiation, having been diagnosed with ADD/ADHD, and symptoms of conduct disorder (Table 4 3). Multinomial lo gistic regression indicated among NH Whites, current daily marijuana use compared to no current use was 18 .00 % more likely with increasing age and 24 .00 % less likely with increasing age of marijuana initiation. Current daily users were also 2.25 times more likely to be male, and 2.34 and 5.39 times more likely to report current tobacco use and lifetime other drug use respectively compared to no n current users Current non daily marijuana users compared to no n current users were 1.33 times more likely to r eport Cs or worse grades in school. Moreover, among NH Whites, both current daily and current non daily marijuana users compared to no n current users were more likely to report current alcohol use (Table 4 5).

PAGE 56

56 Among 505 NH Blacks, 13.07% endorsed current daily marijuana use; 48.51% endorsed current non daily marijuana use and 38.42% endorsed no current marijuana use. Bivariate analyses showed that among NH Blacks, current marijuana use status was significantly associated with age, sex, current tobacco use, current alcohol use, and lifetime other drug use (Table 4 3). Multinomial logistic regression indicated among NH Blacks, compared to no current use, current daily marijuana use was 76 .00 % more likely with increasing age and such users were a lmost four tim es as likely to report current tobacco use. Current non daily marijuana use compared to no current use was 17 .00 % less likely with increasing age; such users were almost as three times more likely to report current alcohol use and 39.00% less likely to rep ort conduct disorder symptoms. Moreover, among NH Blacks, both current daily and current non daily marijuana users compared to no n current users were more likely to be male and report lifetime other drug use (Table 4 5). Among 739 Hispanics, 13.13% endorse d current daily marijuana use; 44.11% endorsed current non daily marijuana use and 42.76% endorsed no current marijuana use. Bivariate analyses showed that among Hispanics, current marijuana use status was significantly associated with sex, family structur e, grades in school, current tobacco use, current alcohol use, lifetime other drug use, having been diagnosed with ADD/ADHD, and symptoms of anxiety and conduct disorder (Table 4 4). Multinomial logistic regression indicated among Hispanics compared to no n current users current daily marijuana users were 2.50 times more likely to be male and 2.87 times more likely to report lifetime use of other drugs. Current daily marijuana use was 26 .00 % less likely with increasing age of marijuana initiation compared to no current use. Current non

PAGE 57

57 daily marijuana us e compared to no current use was 19 .00 % less likely with increasing age; such users were also 36 .00 % less likely to report anxiety symptoms. Moreover, among Hispanics, both current daily and current non dail y marijuana users compared to no n current users were more likely report current tobacco and alcohol use (Table 4 6). Considering the differences by race/ethnicity, we noted that the odds ratios in the separate models overlapped. Further testing revealed th at n o significant interactions were reported be tween race/ethnicity and age, race/ethnicity and sex and race/ethnicity and current alcohol use in predicting current marijuana use 4.5 Discussion The current study investigated the individual level factor s that varied by sex and race/ethnicity in their associations with current marijuana use status among youth who had initiated marijuana use. We first investigated associations by sex. A mong youth who had initiated marijuana use, males were more than twice as likely to endorse current daily marijuana use compared to females (19.28% v ersu s 8.54%). Our finding was consistent with other studies in the literature. Johnston (1981) also found that males were almost twice as likely to endorse daily marijuana use co mpared to females in an older population of high sc hool seniors (Johnston, 1981). Prior studies pointed to a need to examine daily marijuana use in sex specific models. Our analyses found some common individual level factors that predicted current daily marijuana use in both sexes. For instance, current daily marijuana use was more likely among youth who were older and who reported current tobacco and alcohol use and l ifetime use of other drugs among both males and females. Among both, current non daily m arijuana use was also associated with current tobacco use and lifetime other drug use. These findings were in line with previous studies which demonstrated that

PAGE 58

58 mar ijuana use in youth and more so, heavy marijuana use has been asso ciated with higher grade l evel and use of tobacco, alcohol and other substances across both sex es (Schepis et al., 2011; Tu et al., 2008). Current non daily marijuana use was more likely among younger youth across both sex es than no use This might suggest that with increasing age, youth tend to make decisions to either use marijuana daily or to stop entirely They may have tried it, and have decided how it fits or does not fit in their daily routine. Similarly, the age of marijuana initiation was a significant predictor of current daily marijuana use among both sexes. This was in keeping with available study findings that show earlie r age of marijuana initiation was tied to increased use (Sagar et al., 2015). In contrast, some individual level factors significantly predicted current marijuana use among either males or females. Seven specific predictive factors differed in the sex specific models predicting current non daily use with four predicting current daily use. However, while these factors were significantly predictive in one or the other model by sex, the range of values reported in males and females overlapped. Individual variation reported was greater than variation by sex. Further, there was no significant interaction between sex and these factors except self reported ADD/A DHD diagnosis in predicting current marijuana use. Turning first to the male specific model, race/ethnicity and grades in school were significant predictors of both current daily and current non daily marijuana use and mental health problems such as cond uct disorder symptoms and depressive symptoms were significant predictors of current non daily marijuana use among males. Tu and colleagues (2008) found a positive relationship between poor academic performance

PAGE 59

59 and heavy marijuana use among females only; i n contrast, findings from the current study showed this association among males only. In our sample, males who currently used marijuana, but use d it less than 20 days a month, were less likely to have symptoms of either conduct disorder or d epression. Howe ver, Tu and colleagues (2008) reported a significant association between poor self reported mental health among females (Tu et al., 2008), and Schepis and colleagues (2011) reported significant associations between depressed mood and anhedonia and conduct problems across sexes (Schepis e t al., 2011). This wa s one of the first studies to investigate differences among users by current marijuana use status; therefore, these findings should be tested in future studies. Turning to the female specific model, si ngle parent household and having an ADHD diagnosis were both significant predictive factors While statistically different, both males and females showed wide variation in current marijuana use regardless of their family structure. Moreover, among females, having been diagnosed with ADD/ADHD was predictive of current daily marijuana use. Consistent with all sex based differences, the confidence intervals a round this variable overlapped. However, interaction between sex and self reported ADD/ADHD diagnosis s howed the ef fect magnitude was significant among females but not among males This finding was concordant with findings from a previous study (Ottosen et al., 2016) that found higher odds ratio for cannabis abuse or dependence and substance use disorder am ong girls with ADHD compared to boys with ADHD. This suggests that although the prevalence of ADHD is higher among boys than girls, girls with ADHD are at slightly greater risk for marijuana use than boys (Disney et al., 1999; Ottosen et al., 2016).

PAGE 60

60 Secon dly, we investigated associations between individual level factors and current marijuana use status among NH Whites, NH Bla cks, and Hispanics separately. In our sample of youth who had initiated marijuana use, the prevalence of current daily marijuana use was comparable among each race/ethnicity. These findings contrast with those of Johnston that showed Whites as twice as likely as Blacks to report daily marijuana use (Johnston, 1981). These differences in findings may be due to (1) different age groups ( W e studied youth from 13 we re based on high school seniors) ; (2) our study sample only included youth who had which included marijuan a users and non users, and (3) f inally our study findings were reflect changes in marijuana use patterns over time (Johnston, 1981). To date, individual level factors associated with daily or regular marijuana use in youth by each race/ethnicity group hav e not been well studied. The evidence pointed to the potential need to examine daily marijua na use in race specific models. However, our findings showed that while some factors were statistically significantly different by race/ethnicity, there were no sig nificant interactions between race/ethnicity and those significant variables indicating overlapp ing confidence intervals among NH White NH Black and Hispanic estimates Within a particular race, the variation was greater tha n the variation between races. Therefore, we found that individual variation was greater than racial variation. Nevertheless, here we discuss the different model findings by race/ethnicity. Among NH Blacks, NH Whites and Hispanics, current daily marijuana users were more likely to be m ale and to report current use of tobacco and lifetime use of drugs.

PAGE 61

61 Vidourek and colleagues (2017) found that among African American youth, marijuana users were more likely to be male and attend parties with drugs (Vidourek et al., 2017) which is consiste nt with our findings. Ellickson and Morton (1999) had found that use of illicit drugs was higher among Hispanics who had used marijuana compared to White youth (Ellickson and Morton, 1999); we did not find such differences by race/ethnicity. Further, among all three races/ethnicities, with each year increase in age of marijuana initiation, the odds of being a current daily marijuana user decreased by 18% to 26%. Several studies have shown that late age of marijuana initiation was protective for increased le vels of marijuana use (Griffin et al., 2010; Terry 2011; Nelson et al., 2015; Sagar et al., 2015). Our findings were consistent with these studies. Moreover, among NH Blacks and Hispanics, current non daily marijuana use was less lik ely with increasing age and such users were less likely to be males compared to no n current users. This might suggest that among NH Black and Hispanics who had initiated marij uana use, older and male youth a re more likely to either use marijuana daily or t o completely stop using it Current use of alcohol had significant effect s on current daily and current non daily marijuana use among NH Whites and Hispanics only; among NH Blacks, only current non daily marijuana use was associated with current alcohol us e. We also found lower odds for current non daily marijuana use among NH Blacks and Hispanics with conduct disorder and anxiety symptoms, respectively. More studies need to be done to investigate effects of mental health problems and marijuana use status a mong youth who have initiated marijuana use.

PAGE 62

62 Age and current use of alcohol were factors that were specific to two race/ethnicity specific models o ut of three; however, there was no significant interaction by race/ethnicity and these factors in predicting current marijuana use. Also, g iven the fe w differences by race/ethnicity and the high degree of similarity in distributions of factors, these findings should be conservatively interpreted. F uture studies of current marijuana use and its predictors should consider whether it is necessary to differentiate by race/ethnicity. Some limitations of this study should be noted. First, the cross sectional nature of this study did not allow us to infer causal associations between factors and the outcome; instead ou r findings point to important associations. Second, our definition of current daily marijuana use, although in accordance with NSDUH definition, is not ideal. The study did not capture patterns of use among marijuana users prior to the past 30 days. Moreov er, since the responses were categorized in certain blocks, we could not differentiate use on 18 or 19 days with use on 20 or more days which might ha ve caused misclassification in the outcome To ensure sufficient sample size in each outcome group, we cat egorized the outcomes into three groups. Third, since N MAPSS was not designed to look at marijuana use but stimulant use, some other important individual level variables that are associated with youth marijuana use such as socio economic status (SES) dos age and frequency of mar ijuana use, and clinical criteria for cannabis use disorder were not assessed in the study. Regardless of these limitations, N MAPSS was a national study that included a large and diverse sample of youth 10 to 18 years of age and co vered numerous significant factors that are associated with youth substance use. Moreover, it utilized a unique method of sample collection,

PAGE 63

63 entertainment venue intercept that allowed recruitment of youth outside of home or school based settings which migh t have reduced social desirability bias in the study. 4.6 Conclusion To our knowledge, the current study was one of the first studies to examine sex and race/ethnicity specific risk factors associated with current daily marijuana use among marijuana use rs. The current study findings have specific implications for public health professionals for the development of prevention and intervention programs. First, individual level risk factors associated with current daily marijuana use differ by sex and race/e thnicity; however the distributions of factors highly overlap among youth who had initiated use. The most consistent factors that are associated with current daily marijuana use were found to be the use of other substances and early age of marijuana initi ation regardless of sex and race/ethnicity. Future investigations should expand on these findings by using a larger sample size for each sex or racial group to inform prevention or intervention programs.

PAGE 64

6 4 Table 4 1. Socio demographic, behavioral and men tal health characteristics by current daily marijuana use among male and female youth 13 to 18 years of age who had initiated marijuana use in N MAPSS Male (n=1,660; 53.55%) Female (n=1,400; 46.45%) Characteristic No current use (n=617, 37.17%) Current non daily use (n=723, 43.55%) Current daily use (n=320, 19.28%) p value No current use (n=654, 45.42%) Current non daily use (n=663, 46.04%) Current daily use (n=123, 8.54%) p value Sociodemographics Age (MeanSD) 16.341.37 16.261.36 16.76 1.31 .5320 16.311.37 16.161.37 16.591.22 .1339 Race/Ethnicity .0272 .3371 NH White 275 (44.57%) 285 (39.42%) 144 (45.00%) 290 (44.34%) 321 (48.42%) 55 (44.72%) NH Black 92 (14.91%) 146 (20.19%) 53 (16.56%) 102 (15.60%) 99 (14.93%) 13 (10.57%) Hispanic 161 (26.09%) 174 (24.07%) 68 (21.25%) 155 (23.70%) 152 (22.93%) 29 (23.58%) Asian/Asian American 27 (4.38%) 19 (2.63%) 12 (3.75%) 33 (5.05%) 26 (3.92%) 5 (4.07%) Others 62 (10.05%) 99 (13.69%) 43 (13.44%) 74 (11.31%) 65 (9.80%) 21 (17.07%) Family Structure <.0001 <.0001 Living with both parents 296 (47.97%) 330 (45.64%) 110 (34.38%) 326 (49.85%) 263 (40.57%) 37 (30.08%) Living with either mom/dad 245 (39.71%) 310 (42.88%) 140 (43.75%) 239 (36.54%) 30 6 (46.15%) 56 (45.53%) Living with others 76 (12.32%) 83 (11.48%) 70 (21.88%) 89 (13.61%) 88 (13.27%) 30 (24.39%) Grades in school (Cs or worse) 235 (38.09%) 315 (43.57%) 183 (57.19%) <.0001 180 (27.52%) 217 (32.73%) 55 (44.72%) .0005 Behavioral Current tobacco use 120 (19.45%) 165 (22.82%) 174 (54.38%) <.0001 125 (19.11%) 244 (36.80%) 82 (66.67%) <.0001 Current alcohol use 253 (41.00%) 471 (65.15%) 235 (73.44%) <.0001 316 (48.32%) 491 (74.06%) 105(85.37%) <.0001 Lifetime other drug use 180 (29.17%) 262 (36.24%) 243 (75.94%) <.0001 148 (22.63%) 245 (36.95%) 90 (73.17%) <.0001 Age of marijuana initiation (MeanSD) 14.161.87 13.941.94 12.792.32 .0001 14.421.75 14.251.75 12.952.08 .0282 Mental Health ADD/ADHD 1 39 (22.53%) 148 (20.47%) 113 (35.31%) <.0001 84 (12.84%) 115 (17.35%) 38 (30.89%) <.0001 Anxiety symptoms 175 (28.36%) 189 (26.14%) 106 (33.13%) .0695 274 (41.90%) 289 (43.59%) 64 (52.03%) .1148 Conduct disorder symptoms 160 (25.93%) 169 (23.37%) 14 1 (44.06%) <.0001 91 (13.91%) 117 (17.65%) 42 (34.15%) <.0001 Depressive symptoms 163 (26.42%) 160 (22.13%) 90 (28.13%) .0636 256 (39.14%) 269 (40.57%) 64 (52.03%) .0277 Race/Ethnicity Others: Alaskan Native/Eskimo, American Indian, Middle Eastern, Pa cific Islander, and biracial or multiracial; Lived with others: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifetime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants

PAGE 65

65 Table 4 2. Multinomial logistic regression models of sociodemographic, behavioral and mental health characteristics for current marij uana use among male and female youth 13 to 18 years of age who had initiated marijuana use in N MAPSS Model 1 (Male) (n=1,660; 53.75%) Model 2 (Female) (n=1,400; 46.45%) Characteristic Current non daily use versus No current use AOR (95% CI) Current daily use versus No current use AOR (95% CI) Current non daily use versus No current use AO R (95% CI) Current daily use versus No current use AOR (95% CI) Sociodemographics Age 0.90 (0.81, 0.99) 1.21 (1.06, 1.38) 0.85 (0.76, 0.94) 1.33 (1.11, 1.59) Race/Ethnicity NH White Ref Ref Ref Ref NH Black 2.04 (1.44, 2.88) 1.96 ( 1.21, 3.17) 1.25 (0.88, 1.79) 1.17 (0.55, 2.49) Hispanic 1.11 (0.83, 1.49) 0.88 (0.59, 1.33) 0.90 (0.67, 1.21) 0.91 (0.53, 1.59) Asian/Asian American 0.73 (0.38, 1.38) 1.40 (0.61, 3.20) 0.76 (0.42, 1.36) 0.80 (0.26, 2.45) Others 1.68 (1.14, 2.4 6) 1.61 (0.97, 2.67) 0.85 (0.57, 1.26) 1.51 (0.78, 2.90) Family Structure Living with both parents Ref Ref Ref Ref Living with either mom/dad 1.11 (0.87, 1.42) 1.34 (0.95, 2.67) 1.54 (1.19, 1.99) 1.83 (1.11, 3.02) Living with others 0.98 (0.67, 1.43) 1.56 (0.98, 2.48) 1.11 (0.77, 1.60) 1.51 (0.81, 2.81) Self reported school grades (Cs or worse) 1.30 (1.03, 1.66) 1.63 (1.18, 2.24) 1.16 (0.89, 1.50) 1.31 (0.83, 2.07) Behavioral Current tobacco use 1.05 (0.78, 1.42) 2.14 (1.50, 3 .05) 1.85 (1.39, 2.46) 3.02 (1.85, 4.32) Current alcohol use 3.19 (2.50, 4.08) 2.48 (1.76, 3.50) 2.89 (2.26, 3.71) 3.87 (2.19, 6.85) Lifetime other drug use 1.21 (0.93, 1.57) 4.15 (2.92, 5.92) 1.53 (1.15, 2.02) 3.54 (2.14, 5.87) Age of marijuana initiation 0.97 (0.90, 1.04) 0.77 (0.71, 0.84) 1.08 (0.99, 1.17) 0.77 (0.68, 0.87) Mental Health ADD/ADHD 0.91 (0.69, 1.22) 1.15 (0.81, 1.65) 1.37 (0.98, 1.90) 1.83 (1.08, 3.09) Anxiety symptoms 0.90 (0.69, 1.18) 0.80 (0.56, 1.14) 0.99 (0.77, 1. 26) 0.74 7(0.46, 1.19) Conduct disorder symptoms 0.72 (0.55, 0.95) 1.08 (0.77, 1.52) 1.03 (0.74, 1.45) 1.39 (0.82, 2.35) Depressive symptoms 0.75 (0.57, 0.99) 0.79 (0.55, 1.14) 0.89 (0.69, 1.14) 1.08 (0.68, 1.72) Ref = Reference group; Race/Ethnicit y Others: Alaskan Native/Eskimo, American Indian, Middle Eastern, Pacific Islander, and biracial or multiracial; Lived with o thers: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifetime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants

PAGE 66

66 Table 4 3. Socio demographic, behavioral and men tal health characteristics and current daily marijuana use among NH White and NH Black youth 13 to 18 years of age who had initiated marijuana use in N MAPSS NH White (n=1,370; 44.19%) NH Black (n=505; 16.29%) Characteristic No current use (n=565, 41.24%) Current non daily use (n=606, 44.23%) Current daily use (n=199, 14.53%) p value No current use (n=194, 3 8.42%) Current non daily use (n=245, 48.51%) Current daily use (n=66, 13.07%) p value Sociodemographics Age (MeanSD) 16.341.38 16.351.35 16.731.32 .5495 16.431.38 16.231.40 17.260.98 .0022 Male sex 275 (48.67%) 285 (47.03%) 144 (72.36 %) <.0001 92 (47.42%) 146 (59.59%) 53 (80.30%) <.0001 Family Structure .0004 .0550 Living with both parents 314 (55.58%) 312 (51.59%) 87 (43.72%) 59 (30.41%) 65 (26.53%) 12 (18.18%) Living with either mom/dad 195 (34.51%) 229 (37.79%) 71 (35.68%) 96 (49.48%) 145 (59.18%) 37 (56.06%) Living with others 56 (9.91%) 65 (10.73%) 41 (20.60%) 39 (20.10%) 35 (14.29%) 17 (25.76%) Grades in school (Cs or worse) 150 (26.55%) 189 (31.19%) 97 (48.74%) <.0001 69 (35.57%) 101 (41.22%) 32 (48. 48%) .1556 Behavioral Current tobacco use 149 (26.37%) 217 (35.81%) 133 (66.83%) <.0001 15 (7.73%) 36 (14.69%) 24 (36.36%) <.0001 Current alcohol use 292 (51.68%) 472 (77.89%) 161 (80.90%) <.0001 60 (30.93%) 129 (52.65%) 41 (62.12%) <.0001 Lifetime other drug use 162 (28.67%) 224 (36.96%) 165 (82.91%) <.0001 21 (10.82%) 63 (25.71%) 35 (53.03%) <.0001 Age of marijuana initiation (MeanSD) 14.391.79 14.341.75 12.952.22 <.0001 14.411.96 13.932.03 13.022.36 .1680 Mental Health ADD/ADHD 138 (24.42%) 136 (22.44%) 80 (40.20%) <.0001 16 (8.25%) 31 (12.65%) 11 (16.67%) .1305 Anxiety symptoms 190 (33.63%) 217 (35.81%) 73 (36.68%) .6414 63 (32.47%) 86 (35.10%) 21 (31.82%) .7983 Conduct disorder symptoms 93 (16.46%) 95 (15 .68%) 79 (39.70%) <.0001 56 (28.87%) 64 (26.12%) 27 (40.91%) .0634 Depressive symptoms 191 (33.81%) 206 (33.99%) 83 (41.71%) .1024 58 (29.90%) 72 (29.39%) 15 (22.73%) .5110 Lived with others: Lived with foster parents, other relatives or other; Lifet ime other drug use: Lifetime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, s teroids, methamphetamine or inhalants

PAGE 67

67 Table 4 4. Socio demographic, behavioral and men tal health characteristics and current daily marijua na use among Hispanic youth 13 to 18 years of age who h ad initiated marijuana use in N MAPSS Hispanic (n=739; 23.84%) Characteristic No current use (n=316, 42.76%) Current non daily use (n=326, 44.11%) Current daily use (n=97, 13.13%) p value Sociodemographics Age (MeanSD) 16.301.32 16.011.38 16.541.28 .4141 Male sex 161 (50.95%) 174 (53.37%) 68 (70.10%) .0035 Family Structure .0165 Living with both parents 150 (47.47%) 140 (42.94%) 29 (29.90%) Living with either mom/dad 127 (40.19%) 148 (45.40%) 47 (48.45%) Living with others 39 (12.34%) 38 (11.66%) 21 (21.65%) Grades in school (Cs or worse) 130 (41.14%) 149 (45.71%) 60 (61.86%) .0016 Behavioral Current tobacco use 46 (14.56%) 93 (28.53%) 56 (57. 73%) <.0001 Current alcohol use 140 (44.30%) 222 (68.10%) 75 (77.32%) <.0001 Lifetime other drug use 96 (30.38%) 131 (40.18%) 74 (76.29%) <.0001 Age of marijuana initiation (MeanSD) 14.121.80 13.751.97 12.532.28 .0512 Mental Health ADD /ADHD 44 (13.92%) 50 (15.34%) 32 (32.99%) <.0001 Anxiety symptoms 120 (37.97%) 97 (29.75%) 45 (46.39%) .0051 Conduct disorder symptoms 63 (19.94%) 79 (24.23%) 48 (49.48%) <.0001 Depressive symptoms 100 (31.65%) 92 (28.22%) 38 (39.18%) .1191 Lived with others: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifetime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants

PAGE 68

68 Table 4 5 Multinomi al l ogi stic regression models of socio demographic, behavioral and mental health characteristics for current marijuana use among NH White and NH Black youth 13 to 18 years of age who had initiated marijuana use in N MAPSS Model 3 (NH White) Model 4 (NH Bla ck) Characteristic Current non daily use versus No current use AOR (95% CI) Current daily use versus No current use AOR (95% CI) Current non daily use versus No current use AOR (95% CI) Current daily use versus No current use AOR (95% CI) Sociodem ographics Age 0.92 (0.83, 1.02) 1.18 (1.01, 1.38) 0.83 (0.70, 0.99) 1.76 (1.28, 2.41) Male sex 1.03 (0.80, 1.33) 2.25 (1.48, 3.44) 1.70 (1.11, 2.60) 3.76 (1.74, 8.14) Family Structure Living with both parents Ref Ref Ref Ref Livi ng with either mom/dad 1.20 (0.92, 1.56) 0.93 (0.61, 1.42) 1.42 (0.89, 2.26) 1.91 (0.82, 4.44) Living with others 1.17 (0.77, 1.78) 1.32 (0.75, 2.34) 0.84 (0.45, 1.57) 1.38 (0.51, 3.77) Self reported school grades (Cs or worse) 1.33 (1.00,1.77) 1.46 (0.97, 2.19) 1.18 (0.77, 1.79) 1.30 (0.67, 2.50) Behavioral Current tobacco use 1.23 (0.93, 1.63) 2.34 (1.55, 3.54) 1.45 (0.72, 2.93) 3.81(1.57, 9.21) Current alcohol use 3.31 (2.53, 4.32) 2.73 (1.73, 4.32) 2.61 (1.70, 4.01) 1.91 (0.94, 3.88) Lifetime other drug use 1.16 (0.87, 1.54) 5.39 (3.41, 8.52) 2.37 (1.32, 4.24) 4.26 (2.01, 9.03) Age of marijuana initiation 1.01 (0.93, 1.10) 0.76 (0.68, 0.84) 0.99 (0.88, 1.11) 0.82 (0.70, 0.96) Mental Health ADD/ADHD 0.85 (0.64, 1.14) 1.02 (0.67, 1.54) 1.42 (0.71, 2.83) 1.13 (0.42, 3.01) Anxiety symptoms 1.09 (0.83, 1.43) 0.66 (0.43, 1.01) 1.23 (0.78, 1.92) 0.94 (0.45, 1.95) Conduct disorder symptoms 0.80 (0.56, 1.14) 1.17 (0.74, 1.84) 0.61 (0.38, 0.97) 1.19 (0.58, 2.44) Depressive symptoms 0.92 (0.70, 1.20) 1.18 (0.78, 1.78) 0.94 (0.59, 1.49) 0.53 (0.24, 1.20) Ref = Reference group; Lived with others: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifet ime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, methamphetamine or inhalants

PAGE 69

69 Table 4 6. Multinomial logistic regression model of socio demographic, behavioral and mental health characteristics for current marijuana use among Hispanic youth 13 to 18 years of age who had initiated marijuana use in N MAPSS Model 5 (Hispanic) Characteristic Current non daily use versus No current use AOR (95% CI) Current daily use versus No current use AOR (95% CI) Sociodemographics Age 0.81 (0.70, 0.94) 1.24 (0.99, 1.55) Male sex 1.18 (0.83, 1.67) 2.50 (1.37, 4.57) Family Structure Living with both parents Ref Ref Living with either mom/dad 1.23 (0.86, 1.75) 1.80 (0.99, 3.28) Living with others 0.97 (0.56, 1.67) 1.31 (0.59, 2.91) Self reported school grades (Cs or worse) 1.09 (0.78, 1.52) 1.51 (0.88, 2.60) Behavioral Current tobacco use 1.95 (1.25, 3.05) 3.41 (1.86, 6.22) Current alcohol use 2.88 (2.03, 4.11) 2.93 (1.60, 5.37) Lifetime other drug use 1.19 (0.82, 1. 73) 2.87 (1.58, 5.23) Age of marijuana initiation 0.99 (0.89, 1.09) 0.74 (0.64, 0.84) Mental Health ADD/ADHD 1.19 (0.74, 1.92) 1.85 (0.98, 3.52) Anxiety symptoms 0.64 (0.44, 0.93) 0.89 (0.48, 1.64) Conduct disorder symptoms 0.98 (0.64, 1.49) 1.52 (0.84, 2.76) Depressive symptoms 0.80 (0.54, 1.17) 1.15 (0.63, 2.10) Ref = Reference; Lived with others: Lived with foster parents, other relatives or other; Lifetime other drug use: Lifetime us e of cocaine or crack, heroin, club drugs like ecst asy, hallucinogens like LSD or mushrooms, steroid s, methamphetamine or inhalants

PAGE 70

70 CHAPTER 5 INFLUENCE OF PARENTAL AND PEER FACTORS ON CURRENT DAILY MARIJUANA USE AMONG US YOUTH WHO HAD INITIATED MARIJUANA USE 5.1 Background Marijuana is legal for medical and recreational purposes in an increasing number of states in the United States. This legalization may have led to an increased availability of marijuana and decreased perceptions of harm from its regular use especially among adolescents, both of which m ay increase the likelihood of regular marijuana use (Hopfer, 2014; Johnston, 2016). According to the Monitoring the Future (MTF) 2016 survey, 5.4% of the 8 th 14% of the 10 th and 22.5% of the 12 th grade students reported marijuana use in the past month. To develop effective prevention efforts and make greater impact on youth marijuana use, enhancing focus on both parental and peer factors is vital. Parents and peers are the closest social circle of individuals during adolescence and may have important influ substance use (Kandel, 1996; Svensson, 2000). While some research indicated significant associations between parental influence and adolescent substance use (Etz et al., 1998; Svensson, 2000), other research foun d that the impact of parental influence becomes minimal or insignificant after the consideration of peer related factors (Peiser and Heaven, 1996; Bahr et al., 1998). According to the social learning theory, children and adolescents learn their social nor ms through their intimate settings which include their family and friends and they tend to hold similar attitudes and engage in similar behaviors to those in their social settings (Akers and Sellers, 2004; Bahr et al., 2005). For example, in case of substa nce use, adolescents might be exposed to substance use or may acquire positive attitudes towards substance s via parental use of the substance s or parental attitudes towards

PAGE 71

71 drugs. Similarly, they might have higher chances of initiating substance use if the ir friends take such substance either by imitation or reinforcement. However, as adolescents mature they are more likely to have more frequent and strong interactions with friends than their parents. Association with deviant friends at these stages may lea d to detachment with parents and, in turn, result in delinquent behaviors including substance use. Although prior research indicated that parental and peer influences have shown to have significant associations with substance use, these influences have be en defined in different ways across studies, and different variables have been used to concretize these constructs (Hoeve et al., 2009). Parental influence has been measured as parental knowledge/monitoring where parents know where youth were after school or at night or who they are with (Dever et al., 2012; Pinchevsky et al., 2012; White et al., 2006; Yabiku et al., 2010), parental attitudes and perceptions towards drugs (Bahr et al., 2005; King et al., 2012; Musick et al., 2008), parental substance use (K orhonen et al., 2008; Miller et al., 2013; Swift et al., 2008), quality of parent child relationship (Heavyrunner Rioux and Hollist, 2010; Juon et al., 2011; Bahr et al., 2005; Evans et al., 2016), and parental warmth an d control (Mongro Wilson, 2007) Hig h parental knowledge/monitoring (Dever et al., 2012; Pinchevsky et al., 2012; White et al., 2006; Yabiku et al., 2010), negative parental attitudes towards drugs (Bahr et al., 2005; King, Vidourek, & Hoffman, 2012; Musick, Seltzer, & Schwartz, 2008), and p arental warmth and control (Mongro Wilson, 2007) are found to be protective factors against youth marij uana use. Poor family management (Ali et al., 2011; Hampson et al., 2008; Mrug et al., 2010; Perkonigg et al., 2008), parental substance use (Korhonen et al., 2008; Miller

PAGE 72

72 et al., 2013; Swift et al., 2008), and low quality of parent child relationship (Heavyrunner Rioux & Hollist, 2010; Juon et al., 2011; Bahr et al., 2005; Evans et al., 2016) are risk factors for youth marijuana use. To our knowledge, par ental warnings against substance use have not been investigated yet. Similarly, constructs of peer influence that have emerged to be important in predicting youth marijuana use are affiliation with deviant peers (Ali et al., 2011; Hampson et al., 2008; Mru g et al., 2010; Perkonigg et al., 2008) and peer attitude toward /disapproval of substance use (King et al., 2012; Mason et al., 2014), peer marijuana use, (Collins et al., 2011; Walker et al., 2011; Ali et al., 2011; Pinchevsky et al., 2012) and peer subst ance use (Mrug et al., 2010). Studies show that similar to marijuana use, parental and peer influences have also had a significant impact on tobacco and alcohol use among youth. Loke and Mak (2013) found that among adolescents, having authoritative parents (high control, low support) were more predictive of cigarette smoking and having conflicts with parents was more predictive of drinking alcohol whereas having perm issive parents was protective from drinking alcohol. Moreover, having peers who smoked or pe invitation to smoke were both predictive of adolescent smoking and drinking (Loke and Mak, 2013). Youth who took their first sip of alcohol at home, usually parents being the primary source, were more likely to use substance s later (Jackson et al., 20 15). Review studies also reveal the impact of parental monitoring and support, discipline, limiting the availability of alcohol, and parental disapproval of alcohol use (Ryan et al., 2010) and alcohol use. Likewise, having peers who smoked and use of tobacco at home has been associated with

PAGE 73

73 current smoking among adolescents with the former influence reducing and latter remaining static across adolescence (Villanti et al., 2011). Moreover, in our earlier chapter, we found that a substantial proportion of marijuana users reported current tobacco and alcohol use and that current tobacco and alcohol use were significant predictors of current daily marijuana use among youth who had initiated marijuana use (Chapter 3). These findings might suggest a potential mechanism by which current tobacco use and current alcohol use may mediate the association between a number of parental and peer factors and current daily marijuana use. Moreover, a majority of stu dies have focused on the impact of parental and peer influences on marijuana use versu s no use. To our knowledge, the relationship between parental and peer influence and marijuana use among youth who have already initiated marij uana use has been understud ied Therefore, this study aims to test the direct effects of parental and peer factors on current daily marijuana use as well as indirect effects through mediation by current tob acco and alcohol use. I ndividual level predictive factors that were found to be significantly associated with current daily marijuana use in Chapter 3 and are a part of the conceptual model were included in this test Generally, we hypothesized that parental influence ( lack of parental warnings against substances, use of tobacco in the household, having ever gotten alcohol from a parent ) or peer influence (having ever gotten alcohol from a friend and having a higher proportion of close friends who used tobacco or marijuana ) would both directly and indirectly affect current daily m arijuana use through current tobacco and alcohol use. W e also hypothesized that effect of peer influence on current daily marijuana use would be the strongest.

PAGE 74

74 5.2 Methods Data were derived from the National Monitoring of Adolescent Prescription Stimulant s Study (N MAPSS; PI LB Cottler) which consisted of 11,048 youth 10 to 18 years of age. N MAPSS was a cross sectional study that investigated the prevalence of and risk factors for use, misuse and diversion of prescription stimulants among youth. This stud y was conducted from 2008 to 2011 in four waves from ur ban, suburban and rural areas from ten cities across the US. Trained interviewers/recruiters recruited participants from entertainment venues such as shopping malls, movie theatres, parks, arcades, and recreational centers Youth 10 to 12 years of age were given the option to be verbally guided through the survey while youth 13 to 18 years of age completed the survey on their own. The survey took 20 to 30 minutes to be completed. The study was approved by the Washington University in St. Louis Institutional Review Board and the University of Florida Instituti onal Review Board Detailed methods of the study were reported elsewhere (Cottler, Striley, & Lasopa, 2013; Lasopa, Striley, & Cottler, 2015; Strile y, Kelso Chichetto, & Cottler, 2017; Wang, Cottler, & Striley, 2015). 5.2.1 Sample Parental and peer facto rs were only included in Waves Three and F our of N MAPSS (n=5,584). The sample was only restricted to 13 to 18 year olds (n=4,920) due to low rates o f self reported marijuana use among 10 to 12 year olds. Therefore, youth who had initiated marijuana use from W aves Three and Fo ur were included in t he analyses (n=1,586 ). 5.2.2 Measures Outcome variable. The main outcome of interest in this study was cur rent daily marijuana use status among youth who had previously initiated marijuana use.

PAGE 75

75 Participants who reported having ever used marijuana were further asked about their age at first marijuana use and the number of days they used marijuana in the last 30 days. The responses for the latter question were categorized into zero days, one to two days, three to five days, six to nine days, 10 to 19 days, 20 to 29 days and all 30 days. For our analyses, we categorized our main outcome into two groups: No current daily National Survey on Drug Use and Health (NSDUH) study. Predictors. Parental factors included in the analyses were lack of parental warnings against tobacco, alcohol and marij uana use in the past 12 months, having ever gotten alcohol from parents, and use of tobacco in the household. Lack of parental Similarly, lack of parental warnings against alco hol use and marijuana use were sible responses for these parental responses as 0 for having gotten parental warnings sometimes or often and 1 for lack of parental warnings against each substance use. Youth were also asked if they had ever

PAGE 76

76 household sm oke cigarettes or cigars or use chewing t Peer factors. Peer factors included in the analyses were having ever g otten alcohol from a friend, proportion of close friends who used tobacco and proportion of clo s e friends w ho used marijuana. Youth were asked if they ha d ever gotten alcohol from a frien d ( ) The proportion of close friends who used tobacco was calculated using two questions: (1) How many close friends do you have? and (2) How many of your close friends smoke cigarettes or chew tobacco? We calculated the proportion of close friends who used tobacco by dividing the number of close friends who smoked cigarettes or chewed tobacco by the total number of close friends youth had. Similarly, the propor tion of close friends who used marijuana was calculated for each use by dividing the number of close friends, who tried marijuana, even once by the ir total n umber of close friends Individual level factors included in the analyse s were age (13 to 18 year s) sex (male, female) race/ethnicity (Non Hispanic [NH] White, NH Black, Hispanic and other which included Asian/Asian American, Alaskan Native/Eskimo, Pacific Islander, American Indian, Middle Eastern, multiracial or biracial) living with a single par ent or others including relatives or foster parents, self reported grades in school and age of marijuana initiation. Lifetime use of other drugs ( any yes to cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroids, m ethamphetamine or inhalants versus none ) and having been diagnosed with ADD/ADHD were also assessed.

PAGE 77

77 Mediators. We studied two mediators: current tobacco use and current alcohol use. Current tobacco use was determined using two questions. First, youth wer e asked asked they currently smoked cigarettes every day or some days were categorized as current t obacco users. Likewise, current alcohol use was determined using two questions. First, youth were asked categorized as current alcohol users. 5.3 Statistical A nalyses Univariate and bivariate analys es were conducted to report the descriptive statistics of the study sample and to compare characteristics b y current daily and no current daily marijuana use among youth who had initiated use. We then assessed correlations among variables that were include d in the model. The conceptual model was tested us ing path analysis in SAS 9.4 (SAS Institute Inc., Cary, NC, USA). First, a full model w ith all possible paths was tested. Based upon the resu lts of the full path model, paths with no effects were removed an d overall model fit compared The path weights, for both direct and indirect effects of the predictor variables on current daily marijuana use, were calculated. Model fit was assessed using the comparative fit index (CFI) (Bentler, 1990), the Goodness of Fit Index (GFI) and Adjusted Goodness of Fit Index (AGFI) (Joreskog and Sorbom, 1984), and the Root Mean Square Error of Approximation (RMSEA) (Steiger, 1990). Our final parsimonious

PAGE 78

78 model fit the data with a CFI value higher than 0.95, GFI and AGFI value hig her than 0.90 and a RMSEA value lower than 0.05. 5.4 Results Results in Table 5 1 showed th at among a total sample of 1,573 youth who had initiated marijuana use, 14.94 % endors ed current daily marijuana use Overall, the reported mean age was 16.33 years and age of marijuana initiation was14.04 years. Of the sample, 52.70% we re male; and 40.24% were NH White, 55.75% reported living with a single parent or others including foster parents, relatives or others 35.79% reported getting Cs or worse grades in sc hool, 34.52% reported lifetime use of other drugs, and 20.60% reported having been diagnosed with ADD/ADHD. 45.26%, 38.02 % and 33.18 % reported not receiving parental warnings against tobacco, alcohol and marijuana use in the pas t 12 months, respectively; 40.31 % reported use of tob acco in the household; and 30.26% and 73.74 % reported having ever gotten alcohol from a parent and a friend, respectively. A mong the sample, the proportion of close friends who used tobacco was 40 .00 % and the proportion of close f riends who used marijuana was 82 .00 % ; 23.27 % repor ted current use of tobacco and 57.66 % reported current use of alcohol. Bivariate analyses indicated that increasing age, male sex, reporting living with a single parent or others and poor school grades, ear ly age of marijuana initiation, lifetime use of other drugs, self reported ADD/ADHD, use of tobacco in the household, having ever gotten alcohol from a parent, proportion of clo se friends who used tobacco, proportion of close friends who used marijuana and current tobacco and alcohol use were associated with current daily marijuana use. Lack of parental warnings against tobacco, alcohol and marijuana use had no significant impact on current daily marijuana use; however, a higher proportion of no n current d aily marijuana users

PAGE 79

79 reported receiving parental warnings against all three substances compared to current daily u sers. R ace and having ever gotten alcohol from a friend were not significant factors for predicting current daily marijuana use. Table 5 2 sh owed correlations between variables used in the model. Correlation coefficients for all variables were less than 50 .00 %. The tested model explained about 22 .00 % of the variance in current daily marijuana use (Figure 5 1). Below we discussed the direct and indirect effects of the predicting variables. 5.4.1 Direct E ffects The direct effect model examined the direct associations between all parental and peer factors and current daily marijuana use. The standard path coefficients demonstrated that proportion of close friends who used tobacco had a significant effect on current daily m ); however, there were no significant effects of parental factors or any other peer factors on current daily marijuana u ) ) had significant direct impact on cur rent daily marijuana use Similarly, individual level factors, age ( 0.14 ) .09 ) 0.21 ) 0.22 ) ) all had direct effects on current d aily marijuana use (Figure 5 1). 5.4.2 Indirect E ffects The indirect effects tested the mediating effects of current tobacco use and current alcohol use on the associations between parental and peer factors and current daily marijuana use. Individual leve ) ) age of 0.12 ) ) and self reported

PAGE 80

80 ) ; p arental factors: use of t ) and having ever gotte n alcohol from a ); peer factors: having ever gotte n ) and the proportion of close f 0.33) together explained 28.75 % of the variance in current tobacco use. The association between use of tobacco in th e household and current daily marijuana use was fully mediated by current tobacco use. Similarly, the associat ions between having ever gotten alcohol from a parent or a friend and current daily marijuana use were fully mediated by current tobacco use. More over, current tobacco use partially mediated the association between proportion of close friends who used tobacco and current daily marijuana use (Figure 5 1). The i ndividual level factor of ) parental factor of having ever gotten alcohol from a p ) and peer factors of having ever gotten alcohol from a and proportion of close friend explained 26.82 % of the varian ce in current alcohol use. The associat ions between having ever gotten alcohol from a parent or having ever gotten alcohol from a friend and current daily marijuana use were fully mediated by current alcohol use. The association between proportion of close friends who used marijuana and current daily marijuana use was completely m ediated by current alcohol use ( Figure 5 1). 5.5 Discussion Current daily marijuana use has become a serious public health problem. In this study, a signif icant proportion of youth (14.94 %) endorsed current d aily marijuana use Although a number studies have investigated relationships between parental and peer factors and marijuana use, the focus has only been on marijuana users versus no n users. To date, studies have not examined t he impact of parental and peer factors on

PAGE 81

81 current daily marijuana use among youth who had initiated marijuana use. The mediating roles of current tobacco use and current alcohol use in the relationship between different parental and peer factors and curren t daily marijuana use have been understudied. The primary focus of these analyses was to examine the direct and indirect effects of parental and peer factors on current daily marijuana use. Ba sed on our results in Chapter 3 and based on our conceptual mode l, we hypothesized that individual, parental and peer factors would predic t current daily marijuana use. Further, we hypothesized that of these three groups of variables, peer influence would most strongly predic t current daily marijuana use. The total mo del wa s predictive. Several findings we re worth highlighting. First, re search has shown significant associations between tobacco, alcohol and marijuana use among youth (Guxens et al., 2007; Korhonen et al., 2008; Tu et al., 2008; Nelson et al., 2015; ell et al., 2011; Perkonigg et al., 2008; Tu et al., 2008) and ou r findings on youth who had initiated marijuana use were consistent with this body of literature. C urrent tobacco and alco hol use together explained 9 % of variance in current daily marijuana use. In our analyses, although a higher proportion of youth who had initiated marijuana use endorsed current alcohol use than current tobacco use, current tobacco use had a greater effect on current marijuana use. This shows an interesting relationship bet ween tobacco and marijuana use; both are smoked substances and may support and reinforce use of each other. A review paper also concluded that the combined effect s of nicotine and cannabis boost the rewarding effects of each even when consumed in lower do ses (Viveros et al., 2006).

PAGE 82

82 Second, several studies indicated that parental control (Mongro Wilson, 2007) and parental substance use (Korhonen et al., 2008; Miller et al., 2013; Swift et al., 2008) had a significant impact on marijuana use among youth. Ho wever, in our analyses, parental warnings against any of the substances (tobacco, alcohol or marijuana) were not significantly predictive of current daily marijuana use among youth who had already initiated marijuana use. This suggests that once youth have initiated use, their e xperience might contradict what their parents have said about the substance; therefore, those warnings may have been found to be unr eliable or exaggerated and have had no effect on current daily marijuana use. Parental warnings again st all three substances were found to be higher among no n current daily users compared to current daily users in bivariate analyses; however, the effect was insignificant in the conceptual model. Third, other parental factors including use of tobacco by a nyone in the household and having ever gotten alcohol from a parent had no significant direct effects on current daily marijuana use but had indirect effects via current tobacco and/or alcohol use. Prior research found that use of tobacc o in the home was s ignificantly associated with current smoking among youth (Villanti et al., 2011) and youth who sipped alcohol at home and had received alcohol from their parents had higher odds of later substance use (Jackson et al., 2015). As noted, use of tobacco in the household and getting alcohol from a parent we re both indicative of parental permissiveness towards substance use and potentially of easy availability of tobacco and alcohol. This suggests that normalizing substance use by parents and family in the househo ld may be more conseque ntial to the earliest substance use d and less for subsequent drugs used

PAGE 83

83 including marij uana. More over, these measures we re directly relevant to tobacco and alcohol; therefore, they might not have had direct effects on current daily m arijuana use. Fourth, peer factors had the strongest e ffects on current substance use as we hypothesized. The proportion of close friends who used tobacco exerted significant direct and indirect effects via current tobacco use on current daily marijuana us e. Prior studies indicated that having peers who smoked tobacco was significantly associated with marijuana use among youth (Mrug et al., 2010; Villanti et al., 2011; Loke and Mak, 2013); our finding was consistent among youth who had initiated marijuana u se. Studies have also shown having alcohol using peers or deviant peers (Leung et al., 2010) or having peers who smoked (Villanti et al. 2011; Loke and Mak, 2013) was associated with alcohol and tobacco use. In our analyses, having ever gotten alcohol fro m a friend had a significant indirect effect through both current tobacco and alcohol use. This suggests that accessibility of alcohol through friends may increase the likelihood of tobacco and al cohol use which may increase the likelihood of daily mariju ana use. Similarly, the proportion of close peers who used marijuana had a significant indirect effect on current daily marijuana use only via current alcohol use. In our sample, the reported proportion of peer marijuana use was high (82%), reducing variat ion in the sample; this lack of variation may have hampered our ability to distinguish the true influence of peer marijuana use on daily marijuana use. Fifth, the individual level factors such as age, sex, age of marijuana initiation, lifetime use of othe r drugs and self reported ADD/ADHD had direct effects on current daily marijuana use. These f indings were consistent with our earlier findings in Chapter 3.

PAGE 84

84 The current study had some limitations. First, although collected in waves the N MAPSS data was cr oss sectional in nature which limit ed our ability to deduce causality. Second, our outcome variable was categorized into two groups. Youth who used marijuana on zero to 19 days in the past 30 days were all grouped together and remaining youth who used mari juana on 20 or more days in the past 30 days were grouped together. This might have misclassified the outcome variable such that our estimates might have been biased However, our definition of current daily marijuana use is consistent with one of the majo r nationa l studies NSDUH. Third, parental and peer variables that we re used in these analyses were different from the measures used in previous studies making it difficult to compare findings Fourth, peer tobacco and marijuana responses were self reporte d by youth and were not verified by the peers themselves which might have introduced bias Moreover, since N MAPSS was designed to study risk factors for prescription stimulant misuse among youth, some important parental factors such as parental monitoring parental attitudes towards drugs, and parent child relationship and peer factors such as peer attitude toward /disapproval of substance use which were found to be associated with marijuana use among youth in other studies, were not assessed. Even though t here were some limitations, the current study adds to the literature on youth marijuana use in several ways. First, N MAPSS is one of the few national studies with a large and diverse sample of adolescents. It provided a rich source for understanding indi v idual and relationship level risk factors associated with youth marijuana use Moreover, N MAPSS utilized an entertainment venue intercept method for recruitment of youth which enabled it to enroll youth who were absent from school or

PAGE 85

85 home schooled youth. School based national stu dies are more likely to miss these youth The N MAPSS sampling technique might have also reduced social desirability bias because youth did not need parental consent to answer questions When youth answer questions they know their parents or teacher might ask about later they might be more inclined to prevariate A pproval of parents or teachers was not required for youth who chose to complete the interview. 5.6 Conclusion Parental and peer factors are important predictors of curre nt daily marijuana use among youth who had initiated marijuana use. Our findings suggest that parental factors were not directly associated with current daily marijuana use; however, parental factors such as the use of tobacco in the household and receivin g alcohol from a parent had indirect effects on current daily marijuana use through mediation by current tobacco and/or alcohol use. Among the peer factors, the proportion of close peers who used tobacco was directly associated wi th current daily marijuana use. O ther peer factors including receiving alcohol from a friend and proportion of close peers who used marijuana or tobacco had indirect effects on current daily marijuana use through mediation by current tobacco and/or alcohol use. Based on these fi ndings, prevention programs should target aspects of the home environment Parental substance use and accessibility through parents are highly correlated with daily marijuana use among youth. As suggested by social learning theory, parental substance use a nd accessibilit y of substance from parents as well as from peers can have significant impact on youth substance use. Youth may hold similar attitudes and engage in similar behaviors as their parents and /or peers via imitation or reinforcement. While paren tal warnings against substance use may not have an effect

PAGE 86

86 on daily marijuana use among those who have already initiated, parent al monitoring of their children s social network is important. H aving close peers who use substance s is significantly predictive of daily marijuana use. Along with parental and peer factors, individual level factors including older age, being male having an early age of marijuana initiation, using other drugs and poor academic performance are all important risk factors associated with youth marijuana use Each presents a unique target for intervention and on timing To further our understanding of the importance of these factors, f uture research should use longitudinal studies to investigate the role of parental and peer related va riables on daily marijuana use among youth who have already initiated marijuana use.

PAGE 87

87 Table 5 1. Characteristics of youth 13 to 18 years of age who had initiated marijuana u se overall and by current daily and c urrent non daily marijuana use in N MAPSS Wave s Three and F our (n=1,573 ) Characteristics Total (n=1,573) No current daily use (n=1338; 85.06%) Current daily use (n=235, 14.94%) P value Socio demographics Age (Mean SD) 16.33 1.39 16.28 1.39 16.59 1.33 .0017 Male sex 829 (52.70%) 665 ( 49.70%) 164 (69.79%) <.0001 Race/Ethnicity NH White NH Black Hispanic Others 633 (40.24%) 279 (17.74%) 386 (24.54%) 275 (17.48%) 542 (40.51%) 241 (18.01%) 324 (24.22%) 231 (17.26%) 91 (38.72%) 38 (16.17%) 62 (26.38%) 44 (18.72%) .7610 Lived with a single parent or others 877 (55.75%) 719 (53.74%) 158 (67.23%) 0.0001 Grades in School (Cs or worse) 563 (35.79%) 438 (32.81%) 124 (52.77%) <.0001 Age of marijuana initiation (Mean SD) 14.04 2.00 14.27 1.83 12.74 2.36 <.0001 Lifetime use of other drugs 543 (34.52%) 367 (27.43%) 176 (74.89%) <.0001 ADD/ADHD 324 (20.60%) 241(18.01%) 83 (35.32%) <.0001 Parental factors Lack of parental warnings against tobacco use 712 (45.26%) 595 (44.47%) 117 (49.79%) .1309 Lack of parental warnings against alcohol use 598 (38.02%) 496 (37.07%) 102 (43.40%) .0651 Lack of parental warnings against marijuana use 522 (33.18%) 446 (33.33%) 76 (32.34%) .7656 Parental warnings against all three substances 588 (37.38%) 509 (38.04%) 79 (33.62%) .0099 Use of tobacco in the household 634 (40.31%) 520 (38.86%) 114 (48.51%) .0054 Ever got alcohol from parent 476 (30.26%) 389 (29.07%) 87 (37.02%) .0139 Peer factors Ever got alcohol from friend 1160 (73.74%) 976 (72.94%) 184 (78 .30%) .0854 Proportion of close friends who used tobacco 0.40 0.38 0.36 0.37 0.64 0.38 <.0001 Proportion of close friends who used marijuana 0.82 0.28 0.80 0.29 0.92 0.22 <.0001 Current tobacco use 366 (23.27%) 248 (18.54%) 118 (50.21%) <.0001 Current alcohol use 907 (57.66%) 734 (54.86%) 173 (73.62%) <.0001 Race/Ethnicity Others: Asian/Asian American, Alaskan Native/Eskimo, American Indian, Middle Eastern, Pacific Islander, and biracial or multiracial; Lifetime other drug use: Lifetime use of cocaine or crack, heroin, club drugs like ecstasy, hallucinogens like LSD or mushrooms, steroi ds, methamphetamine or inhalants

PAGE 88

88 Table 5 2. Correlation coefficients between variables included in the model P p <.05; ** p<.001; *** p<.0001 Characteristics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1. Age -.04 .01 .02 .41*** 0.05 .00 .08* .07* .16*** .04 .07 .15*** .07* .08** .07* .14*** .08* 2. Sex -.00 .16*** .12*** .09** .07* .02 .03 .05 .04 .02 .12*** .01 .00 .03 .05* .14*** 3. Living with either a single par ent or others -.12*** .11*** 07* .05* .05* .04 .01 .08* .01 6* .05* .08* .05* .03 .10** 4. Grades in School (Cs or worse) -.17*** .13*** .12*** .02 .04 .00 .04 .04 .02 .15*** .09** .12*** .00 .15*** 5. Age of marijuana initiation .27*** .15*** .01 .01 .03 .10*** .05* .02 .21*** .12*** .22*** .03 .27*** 6. Lifetime use of other drugs -.11*** .01 .03 .02 .06* .12*** .22*** .27*** .18*** .32*** .26*** .36*** 7 Having been diagnosed with ADD/ADHD -.00 .0 4 .00 .02 .08* .02 .15*** .10*** .20*** .03 .15*** 8 Lack of parental warnings against tobacco use -.36*** .37*** .07* .02 .02 .01 .01 .01 .06* 0.04 9 Lack of parental warnings against alcohol use -.40*** .03 .14*** .04 .03 .02 .05 .03 .05 10 Lack of parental warnings against marijuana use -.02 .04 .01 .00 .01 .04 .01 .01 11 Use of tobacco in the household -.07* .07* .17*** .06* .19*** .06* .07* 12 Ever gotten alcohol from a parent -.21*** .09** .05* .14*** .21*** .06* 13 Ever gotten alcohol from a friend -.17*** .20*** .18*** .47*** .04 14 Proportion of close friends who used tobacco -.36*** .45*** .19*** .27*** 1 5 Proportion of close friends who us ed marijuana -.15*** .21*** .15*** 1 6 Current tobacco use -.27*** .27*** 1 7 Current alcohol use -.14*** 1 8 Current daily marijuana use -

PAGE 89

89 Note: Only significant paths were included in the figure, standardized coefficients were reported; *= p <0.05 Fig ure 5 1. Direct and indirect effects of individual, parental and peer factors on current daily marijuana use among youth 13 to 18 y ears of age who had initiated marijuana use (n=1,573)

PAGE 90

90 CHAPTER 6 CONCLUSION Marijuana is the most commonly used illicit drug by youth in the United States About 6.5% o f 12 to 17 year olds reported marijuana use i n the past 30 days (SAMHSA, 2017 ). Among middle and high school stude nts, the prev alence estimates of past 30 day marijuana use were 5.4% 14%, and 22.5% among 8 th 10 th and 12 th grade students indicating an increase by grade level (Johnston et al., 2016). Research indicated that the prevalence has decreased among 8 th and 10 th grade stu dents and remained static among 12 th grade students in 2016 compared to 2011. However, with the legalization of marijuana for medical purposes and adult recreational use in a number of states, views and attitud es towards marijuana among youth may be changi ng According to the Monitoring the Future 2016 survey, almost a half of high school seniors used marijuana once in their lifetime and there was a significant decrease in the proportion of youth who perceived regular marijuana use as detrimental (Johnston et al., 2016). Marijuana use is important to study because it can cause significant adverse developmental, neurobiological, psychosocial and behavioral consequences for youth (Volkow et al., 2014; Hall and Degenhardt, 2014 ). Current evidence has shown cons istent adverse effects of regular marijuana use during adolescence. In most epidemiological studies, regular marijuana use is measured as daily or almost daily use (Hall & Degenhardt 2009). It has been reported that 10% of individuals who ever used mariju ana become daily marijuana users (Hall & Degenhardt, 2014). Several studies have revealed a number of risk and protective factors associated with youth marijuana use; much less research has been done on the risks for daily marijuana use. Since daily mariju ana use among youth is of serious concern, it is urgent to address the problem of

PAGE 91

91 current daily marijuana use among youth who have already initiated use and the ways in which daily users are different from non daily users Compounding the problem of a lack of res earch is the lack of consistency in definitions of daily marijuana use across studies. To make the findings comparable, we defined daily marijuana use as the use of marijuana on 20 or more days in the past 30 days, mirroring the National Survey on D rug Use and Health (NSDUH). To develop strategic prevention programs, understanding the factors associated with daily marijuana use among youth is vital. Therefore, this dissertation intended to characterize the factors associated with current daily mariju ana use among youth wh o had initiated marijuana use. First, individual level factors that predict current daily marijuana use among youth were investigated using multinomial logistic regression. Second, the characteristics of youth who endorsed current dai ly marijuana use were examined by sex and race/ethnicity. Lastly, the direct and indirect effects of parental and peer factors on current daily marijuana use were studied using path analysis A ll analyses used data from the National M onitorin g of Adolescen t Prescription Stimulant s S tudy (N MAPSS ) 6.1 Main Findings Results from the first analysis (Chapter 3) showed that 14.29% of youth who had initiated marijuana use endorsed current daily marijuana use. Overall, our sample showed that youth who endorsed c urrent daily marijuana use reported high rates of poor academic performance, current tobacco and alcohol use, lifetime use of drugs and mental health problems including ADD/ADHD and symptoms of conduct disorder. Multinomial logistic regression indicated th at behavioral factors such as current tobacco use, current alcohol use, lifetime drug use and early age of marijuana initiation had strong associations with current daily marijuana use. Current daily marijuana users were

PAGE 92

92 also more likely to be older, male and NH Black or individual of other race s (Alaskan Native/Eskimo, Pacific Islander, American Indian, Middle Eastern, multiracial or biracial) compared to NH White, and were more likely to report living with a single parent or with foster parents, re latives or others and getting Cs or worse grades in school. Having been diagnosed with ADD/ADHD also independently increased the likelihood of current daily marijuana use. Mental health problems including symptoms of anxiety, conduct disorder, and depression whi ch were hypothesized to be significant predictors of current dai ly marijuana use, were not. Our findings are unique because our study sampl e consisted of only those youth who had initiated marijuana use and comparison was done between current daily m arijua na use and no current use. H owever, we found similar risk fac tors for daily marijuana use in multiple studies that compared marijuana use to no use. Therefore, p revention efforts should address daily marijuana use among youth accounting for these identifie d risk factors. Targeting polysubstance users and d elaying marijuana initiation may reduce the risk of becoming a daily marijuana user and the adverse consequences associated with daily marijuana use. Moreover, older youth should get booster intervention f or subsequent decrease s in daily marijuana use. In our second set of analyses, we stratified youth who had initiated marijuana use by sex and race/ethnicity to describe the characteristics of current daily marijuana users separately. This enabled us to exa mine factors associated with current daily marijuana use unique to a particular sex and to a specific race/ethnicity This specification might have provided useful information to help design targeted prevention efforts. We found that the prevalence s of cur rent daily marijuana use among males and females were 19.25% and 8.54%, respectively in our sample. Multinomial logistic regression

PAGE 93

93 indicated that current daily marijuana use was associated with increasing age, current tobacco and alcohol use, lifetime ot her drug use and early age of marijuana initiation across both sex es These findings were in line with our hypotheses. Moreover, among males, being NH Black and reporting Cs or worse grades in school were significant predictors of current daily marijuana u se; among females, living with a single parent and having been diagnosed with ADD/ ADHD were significant predictors of current daily marijuana use. However, we found that the effect magnitudes of these factors showed very similar and overlapping ranges betw een the sexes except for self reported ADD/ADHD diagnosis. Further testing revealed the model differences were not significant except that c urrent daily marijuana users were more likely to report having been diagnosed with ADD/ADHD in the female only mode l This effect was not significant among males. F indings from these analyses show that it may not be important to distinguis h risk factors between the sexes except to target women with ADD/ADHD who may be at an additional risk for daily marijuana use Nex t, we studied the influence of individual level factors on current daily marijuana use among NH Whites, NH Blacks, and Hispanics. We found that the prevalence estimates of current daily marijuana use among NH White s, NH Blacks, and Hispanics were comparabl e. Multinomial logistic regression indicated that current daily marijuana use was associated with being male, reporting current tobacco use and lifetime other drug use and early age of marijuana initiation among NH Whites, NH Blacks, and Hispanics. These f indings were consistent with our hypotheses, except conduct disorder symptoms were insignificant. Current daily marijuana use was associated with current alcohol use among NH Whites, and Hispanics and with

PAGE 94

94 increasing age among NH Whites and NH Blacks only. When tested for interaction between race/ethnicity and factors that were significant in either one or the other model (age and current alcohol use), we did not find any significant differences between race specific models From these findings, it follows that it may not be important to distinguis h individual level risk factors between the races H owever we did not measure other factors such as culture, percept ion and attitudes towards drugs which might be important in race/ethnicity specific models Futur e investigations should expand on these finding s by using a larger sample for each racial group to better inform prevention or intervention programs. Along with individual level factors, there have also been reports of parental and peer factors that predic ted marijuana use among youth. Thus, we aimed to examine direct and indirect effects (via current tobacco and alcohol use) on current daily marijuana use among youth who had already in itiated marijuana use. No non users were included. Using path analysis we found that individual level factors of increasing age, male sex, early age of marijuana initiation, lifetime use of illicit drugs and having been diagnosed with ADD/ADHD significantly predicted current daily marijuana use. Among the parental and peer fa ctors, only the proportion of close friends who used tobacco had a direct effect on current daily marijuana use. A parental factor, use of tobacco in the household, increased the likelihood of current daily marijuana use by the mediating the effect of curr ent tobacco use. Another parental factor, having ever gotten alcohol from a parent, increased the likelihood of current daily marijuana use by the mediating effects of both current tobacco use and current alcohol use. However, parental warnings against tob acco, alcohol or marijuana use did not have significant any

PAGE 95

95 effect on current daily marijuana use. Among peer factors, both current tobacco and /or current alcohol use mediated the relationship between ever getting alcohol from a friend and current daily ma rijuana use whereas only current alcohol use mediated the relationship between proportion of close friends who used marijuana and current daily marijuana use. Moreover, the proportion of close friends who used tobacco also increased the likelihood of curre nt daily marijuana use b y the mediating effect of current tob acco use Most associations between parental and peer factors and current daily marijuana use were fully or partially mediated by current tobacco and current alcohol use. Most who used marijuana daily also endorsed current alcohol and tobacco use. Since these peer and parental factors affected marijuana use through alcohol and tobacco use, and since most of the youth who used marijuana used one or two of the other drugs, use of any or all are impo rtant to prevention efforts. To prevent current daily marijuana use, efforts should target current tobacco and alcohol use and the factors that directly influence the use of these substances, including parental and peer factors. Prevention programs should focus on parental and peer substance use as well as accessibility of substances from parents and peers to decrease the rates of daily marijuana use among youth. Trainings on changing parental and peer behavior, family management practices (communication, e ngagement, and supportiveness) and reducing daily marijuana use among youth. Moreover, teaching drug resistance skills to youth so that they can refuse substance off ers ma de by parents or peers show promise Overall, this dissertation has provided tests of the influence of individual and relationship level factors on current daily marijuana use by identifying these factors

PAGE 96

96 among teens who had initiated marijuana use. As no ted, sociodemographic factors such as male sex and being older and behavioral factors such as current use of tobacco and alcohol, use of other drugs and early age of marijuana initiation were important factors that predicted current daily marijuana use. Mo reover, relationship level factors such as tobacco use in the household, acquiring alcohol from parents or friends, and having many close peers who use tobacco or marijuana were also found to predict current daily marijuana use. It has also provided eviden ce for the development of preventive efforts to addr ess current daily marijuana use. Since peer factors had stronger effects than parental factors on current substance use, parents should focus on monitoring their children intervening to limit the influence of substanc e using peers. This will likely be more effective than warning them against the use of substances. 6.2 Strengths and Limitations This dissertation presents one of the first studies to investigate factors associated with curr ent daily marijuana use among teens who had already initiated marijuana use It utilized a national cross sectional study, the National Monitoring of Adolescent Prescription Stimulant s Study (N MAPSS) that consisted of a large and diverse sample of youth a nd a wide range of factors associated with substance use. N MAPSS used a n entertainment venue intercept method recruiting youth from shopping malls, parks, recreational centers, mov ie theatres, arcades, libraries and skate parks. This method enabled the s tudy team to capture youth away from home or school based settings This may have reduced social desirability bias in our findings. Even though this dissertation had several strengths, some important limitations should be noted. First, N MAPSS data were cr oss sectional in na ture; therefore, this limit s the ability to infer causality between the tested risk factors and outcome variable s

PAGE 97

97 Second, data were measured through self report. However, self reported measures of substance use have been found to be hig hly reliable, about 95% as reliable as biochemical assessments in one study (Hjorthoj et al., 2012). Th ird, although the definition of current daily marijuana use was consistent with the NSDU H study, this definition is less than ideal. Y outh who initiated marijuana use were grouped into three categories in the first two studies and into two groups in the third study; we may have misclassified t he outcome variable in some cases However, use of this definition will allow comparison between these findings and the NSDUH. Fourth, the methodology of the study may also have limited findings on parental and peer influence that other meth ods might have captured. Social network analysis directly gathers information from peers (Bauman and Ennett, 1996; Ennett et al., 2006); this method might have provided a more reliable measure of peer influence. Peer and parental direct reports would also be more likely to capture influence (Griffin et al., 2000; Martins et al, 2008) Lastly, measures of frequency, quantity, cannabis use disorder, parental substance use, parental and peer attitudes towards marijuana, parental monitoring, and quality of the parent child relationship were not assessed, all of which might have been important factors for predicting youth marijuana use in our sample. 6.3 Future Research With an increasing number of states in the US changing marijuana laws to allow medical and/or adult recreational use in recent years marijuana use continues to rise. We found a number of factors that influence current daily marijuana use among youth who had initiated marijuana use. Future research should assess the associations between individual and relationship level factors and current daily marijuana use among youth marijuana initiators using other ann ual national studi es Findings can then

PAGE 98

98 be compared Moreover, longitudinal studies should assess the direction of possible pathways linking the factors and current daily marijuana use to assess causation. Our findings show ed significant associations between current daily m arijuana use and sociodemographic factors including age and sex and behavioral factors including current use of tobacco and alcohol, use of other drugs and age of marijuana initiation. Mental health problems did not significantly predict current daily mari juana use, except among female youth H aving been diagnosed with ADD/ADHD was associated with current daily marijuana use among girls More studies should be conducted to study the dynamics and directionality of mental health problems and current marijuana use status. Many studies have found directions both from and to mental health problems from substance use. A number of small studies are using cannabinoids to treat mental distress and specifically, ADHD (Cooper et al., 2017). Parental and peer factors th at influence current daily marijuana use should also be examined using longitudinal studies. This dissertation has not considered the consequences of daily use, but rath er risk factors for daily use. Intervention programs should consider risk factors in o rder to develop prevention strategies against current daily marijuana use and its consequences. The intervention programs that have been implemented targeting identified risk factors should be subject to further rigorous evaluation and replication This re mains a highly relevant and active area of research. Finally, it will be important to extend this research to explore the health outcomes and other negative consequences of daily marijuana use. This remains a significant public health challenge.

PAGE 99

99 APPENDIX NATIONAL MONITORING O F ADOLESCENT PRESCRIPTION STIMULANT S S TUDY Table A 1. N MAPSS questions used in analyses Characteristics Questions in N MAPSS Responses Age Ques002. How old are you? Discrete Ranges from 10 to 18 Sex Ques001. What is your gender? Dichotomous 1. Male 2. Female Race/ethnicity Ques003. Are you Hispanic or Latino? Ques004. Which racial or ethnic group best describes you? Dichotomous 1. No 2. Yes Categorical 1. Alaskan Native/Eskimo 2. American Indian 3. Asian or Asian American 4. Black or African American 5. Middle Eastern 6. Pacific Islander 7. White, Euro American, Caucasian 8. Biracial or Multiracial 9. Other Family structure Ques006. In the last 7 days, who have you lived with? Categorical 1. Lived with mom and dad at the s ame time 2. Lived with mom and dad, but not at the same time 3. Lived with mom only 4. Lived with dad only 5. Lived with foster parents 6. Lived with other relatives 7. Lived with others Grades in school Ques013. Which of the following best describes you r grades? Categorical know Self reported general health Ques025. Is your general health excellent, good, fair, or poor? Categorical 1. Excellent 2. Good 3. Fair 4. Poor Current tobacco use Ques120. Have you ever smoked? Ques122. Do you still smoke cigarettes every day or some days? Dichotomous 1. No 2. Yes Categorical 1. No 2. Yes, I currently smoke cigarettes everyday 3. Yes, I currently smoke cigarettes som e days.

PAGE 100

100 Table A 1. Continued Characteristics Questions in N MAPSS Responses Current alcohol use Ques124. Have you ever had a beer, a glass of wine, or any other alcoholic drink, not just a sip? Ques127. In the last 30 days, on how many days did you drin k alcohol? Dichotomous 1. No 2. Yes Categorical 1. 0 2. 1 or 2 3.3 to 5 days 4. 6 to 9 days 5. 10 to 19 days 6. 20 to 29 days 7. All 30 days Current marijuana use Ques114. Have you ever used marijuana? Ques116. In the last 30 days, how many days did y ou use marijuana? Dichotomous 1. No 2. Yes Categorical 1. 0 2. 1 or 2 3.3 to 5 days 4. 6 to 9 days 5. 10 to 19 days 6. 20 to 29 days 7. All 30 days Age of marijuana initiation Ques115. At what age did you first use marijuana? Continuous Lifetime other d rug use a. Cocaine or crack b. Heroin c. Club drugs like ecstasy d. Hallucinogens like LSD or mushrooms e. Anabolic steroids f. Cough syrup/purple drank to get high g. Methamphetamine h. Inhalants like gasoline or paint? Dich otomous for each drug 1. No 2. Yes ADD/ADHD Ques030. Has a doctor ever told you or your parents that you have Attention Deficit Disorder (ADD) or Attention Deficit Hyperactivity Disorder (ADHD)? Categorical Anxiety symptoms Ques036. Have you ever felt worried or stressed for 6 months or more? Dichotomous 1. No 2. Yes Conduct Disorder symptoms Ques031. Have you ever gotten into a lot of trouble at home or at school or ran away from home overnight? Ques032. Have you ever use d or threatened someone with a weapon? Dichotomous 1. No 2. Yes Dichotomous 1. No 2. Yes Depressive symptoms Ques035a. In the last 12 months, have you had 2 weeks or more when you lost interest in things? Ques035b. In the last 12 months, have you had 2 weeks or more when you felt down or depressed? Dichotomous 1. No 2. Yes Dichotomous 1. No 2. Yes

PAGE 101

101 Table A 1. Continued Characteristics Questions in N MAPSS Responses Parental Influence Ques118a. In the past 12 months, how often has a parent or guardian warned you not to use marijuana? Ques123a. Does anyone in your household smoke cigarettes or cigars or used chewing tobacco, snuff, or dip? Ques123b. In the past 12 months, how often has a parent or guardian warned you not to smoke or chew tobacco? Qu es129. Have you ever gotten alcohol from your parent? Ques129h. In the past 12 months, how often has a parent or guardian warned you not to use alcohol? Categorical 1. Never 2. Sometime 3. Often Dichotomous 1. No 2. Yes Categorical 1. Never 2. Someti me 3. Often Dichotomous 1. No 2. Yes Categorical 1. Never 2. Sometime 3. Often Peer influence Ques129e. Have you ever gotten alcohol from your friend? Ques134. How many close friends do you have? Ques135a. How many of your close friends have tried m arijuana, even once? Ques135b. How many of your close friends smoke cigarettes or chew tobacco? Dichotomous 1. No 2. Yes Discrete Ranges from 0 to more than 10 Discrete Ranges from 0 to more than 10 Discrete Ranges from 0 to more than 10

PAGE 102

102 LIST OF REFERENCES Adlaf, E.M., Paglia, A., 2003. Drug use among Ontario students, 1977 2003: Findings from OSDUS. Toronto, Ontario: Center for Addiction and Mental Health. Akers, R. L., Krohn, M. D., Lanza Kaduce, L., & Radosevich, M., 1979. Social learning and d eviant behavior: A specific test of a general theory. Am. Sociol. Rev. 44, 636 655. Ali, M. M., Amialchuk, A., Dwyer, D. S., 2011. The Social contagion effect of marijuana use among adolescents. PLoS One 6 e16183. American Civil Liberties Union, 2013. The War on Marijuana in Black and White. Azofeifa, A., 2016. National estimates of marijuana use and related indicators National Survey on Drug Use a nd Health, United States, 2002 2014. MMWR. Surveill. Summ 65 Bahr, S. J., Hoffmann, J. P., Yang, X., 2005. Parental and peer influences on the risk of adolescent drug use. J. Prim Prev 26 529 551. Bahr, S. J., Maughan, S. L., Marcos, A. C., Li, B., 1998. Family, religion, and the risk of adolescent drug use. J. Marriage Fam. 60, 979 992. Barkley, R.A., Fischer, M., Smallish, L., Fletcher, K., 2004. Young adult follow up of hyperactive children: antisocial activities and drug use. J. Child Psychol. P sychiatry. 45, 195 211. Bauman, K. E., Ennett, S. T., 1996. On the importance of peer influence for adolescent drug use: commonly neglected considertations. Addiction. 91, 185 198. Bentler, P. M., 1990. Fit indices, LaGrange multipliers, constraint chang es, and incomplete data in structural models. Multivariate Behav. Res. 25, 163e172. Collins, D., Abadi, M.H., Johnson, K., Shamblen, S., Thompson, K., 2011. Non medical use of prescription drugs among youth in an Appalachian population: Prevalence, predic tors, and implications for prevention. J. Drug Educ. 41, 309 326. Conroy, D. A., Kurth, M. E., Strong, D. R., Brower, K. J., Stein, M. D., 2016. Marijuana use patterns and sleep among community based young adults. J. Addict. Dis 35 135 143. Cooper, R. E., William, E., Seegobin, S., Tye, C., Asherson, P., 2017. Cannabinoids in attention deficit/hyperactivity disorder: A randomized control trial. Eur. Neuropsychopharmacol. 27, 795 808.

PAGE 103

103 Cottler, L. B., Striley, C. W., Lasopa, S. O., 2013. Assessing prescr iption stimulant use, misuse, and diversion among youth 10 18 years of age. Curr. Opin. Psychiatry. 26 511 519. Creemers, H. E., Dijkstra, J. K., Vollebergh, W. A. M., Ormel, J., Verhulst, F. C., Huizink, A. C., 2010. Predicting life time and regular can nabis use during adolescence; the roles of temperament and peer substance use: the TRAILS study. Addiction. 105, 699 708. Davis, C. E., Hyde, J. E., Bangdiwala, S.I., Nelson, J.J., 1986. An example of dependencies among variables in a conditional logisti c regression, in: Modern Statistical Methods in Chronic Disease Epidemiology, 140 147. Bachman, J. G., 2012. Predicting risk taking with and without substance use: the effe cts of parental monitoring, school bonding, and sports participation. Prev. Sci 13 605 615. Dinieri, J. A., & Hurd, Y. L., 2012. Rat models of prenatal and adolescent cannabis exposure. Methods in Molecular Biology (Clifton, N.J.) 829 231 242. Disney E.R., Elkins, I.J., McGue, M., Iacono, W.G., 1999. Effects of ADHD, conduct disorder, and gender on substance use and abuse in adolescence. Am. J. Psychiatry. 156, 1515 1521. Ellickson, P. L., Morton, S. C., 1999. Identifying adolescents at risk for har d drug use: Racial/ethnic variations. J. Adolesc Health 25 382 395. Ennett, S. T., Bauman, K. E., Hussong, A., Faris, R., Foshee, V.A., Cai, Li, DuRant, R. H., 2006. The peer context of adolescent substance use: Findings from social network analysis. J. Res Adolesc. 16, 159 186. Etz, K. E., Robertson, E. M., Ashery, R. S., 1998. Drug abuse prevention through family based interventions: Future research. In R. S. Ashery, E. B. Robertson, K. L. Kumpfer (Eds.), Drug abuse prevention through family intervent ions (pp. 1 11). NIDA Research Monograph 1777. Rockville, MD: National Institute on Drug Abuse. Evans, C. B., Cotter, K. L., Rose, R. A., Smokowski, P. R., 2016. Substance use in rural adolescents: The impact of social capital, anti social capital, and so cial capital deprivation. J. Addict Dis. 1 14. Farhat, T., Simons Morton, B., Luk, J. W., 2011. Psychosocial correlates of adolescent marijuana use: variations by status of marijuana use. Addict. Behav. 36 404 407.

PAGE 104

104 Gogtay, N., Giedd, J. N., Lusk, L., H ayashi, K. M., Greenstein, D., Vaituzis, A. C., Thompson, P. M., 2004). Dynamic mapping of human cortical development during childhood through early adulthood. Proc. Natl. Acad. Sci. U.S.A. 10 1, 8174 8179. Griffin, K. W., Botvin, G. J., Scheier, L. M., Di az, T., Miller, N. L., 2000. Parenting practices as predictors of substance use, delinquency, and aggression among urban minority youth: Moderating effects of family structure and gender. Psycho. Addict Behav. 14, 174 184. Griffin, K. W., Botvin, G. J., S cheier, L. M., & Nichols, T. R., 2002. Factors associated with regular marijuana use among high school students: a long term follow up study. Subst Use Misuse 37 225 238. Griffin, K. W., Bang, H., Botvin, G. J., 2010. Age of alcohol and marijuana use on set predicts weekly substance use and related psychosocial problems during young adulthood. J. Subst Use. 15, 174 183. Guxens, M., Nebot, M., Ariza, C., 2007. Age and sex differences in factors associated with the onset of cannabis use: a cohort study. Dr ug Alcohol Depend 88 234 243. Haberstick, B. C., Young, S. E., Zeiger, J. S., Lessem, J. M., Hewitt, J. K., Hopfer, C. J., 2014. Prevalence and correlates of alcohol and cannabis use disorders in the United States: Results from the national longitudina l study of adolescent health. Drug Alcohol Depend. 136 158 161. Hall, W. Degenhardt, L ., 2007. Prevalence and correlates of cannabis use in developed and developing countries. Curr. Opin. Psychiatry 20, 393 397. Hall, W., Degenhardt, L., 2009. Advers e health effects of non medical cannabis use. Lancet 374 1383 1391. Hall, W., Degenhardt, L., 2014. The adverse health effects of chronic cannabis use. Drug Test Anal. 6, 39 45. Hampson, S. E., Andrews, J. A., Barckley, M., 2008. Childhood p redictors o f adolescent marijuana use: Early sensation seeking, deviant peer affiliation, and social i mages. Addict Behav 33 1140 1147. T., 2015. Medical marijuana laws and adolescent marijuana use in the USA from 1991 to 2014: results from annual, repeated cross sectional surveys. Lancet Psychiatry 2 601 608.

PAGE 105

105 Hayaki, J., Hagerty, C. E., Herman, D. S., de Dios, M. A., Anderson, B. J., Stein, M. D. (2010). Expectancies and marijuana use frequency and severity among young females. Addict. Behav 35 995 1000. Heavyrunner Rioux, A. R., & Hollist, D. R., 2010. Community, family, and peer influences on alcohol, marijuana, and illicit drug use among a sample of Native American youth: an analysis of predictive factors. J. Ethn. Subst Abuse 9 260 283. Hjorthoj, C. R., Hjorthoj, A. R., Nordentoft, M., 2012. Validity of timeline follow back for self reported use of cannabis and other illicit substances: systematic review and meta analysis. Addict. Behav. 37, 225 33. Hoeve, M., Dubas, J. S., Eichelsheim, V. I., van der Laan, P. H., Smeenk, W., Gerris, J. R. M., 2009. The relationship between parenting and delinquency: A meta analysis. J. Abnor. Child Psychol 37 749 775. Homel, J., Thompson, K., Leadbeater, B., 2014. Trajectories of marijuana use in youth ages 15 25: Implications for postsecondary education experiences. J. Stud. Alcohol Drugs 75 674 683. Hopfer, C., 2014. Implications of marijuana legalization for adolescent s ubstance use. Subst Abus.35, 331 335. Horton, J., Compton, W., Cottler, L.B., 2000. Reliability of substance use disorder diagnoses among African Americans and Caucasians. Drug and Alcohol Dependence. 57, 203 209. Horwood, L. J., Fergusson, D. M., Hayatb akhsh, M. R., Najman, J. M., Coffey, C., Patton, G. C., Hutchinson, D. M., 2010. Cannabis use and educational achievement: findings from three Australasian cohort studies. Drug Alcohol Depend. 110 247 253. Jackson, K. M., Barnett, N.P., Colby, S. M., Ro gers, M. L., 2015. The prospective association between sipping alcohol by the sixth grade and later substance use. J. Stud. Alcohol Drugs. 76, 212 221. Johnson, R. M., Fairman, B., Gilreath, T., Xuan, Z., Rothman, E. F., Parnham, T., Furr Holden, C. D. M. 2015. Past 15 year trends in adolescent marijuana use: Differences by race/ethnicity and sex. Drug Alcohol Depend. 155 8 15. Johnston, L. D., 1981. Characteristics of the daily marijuana user. The University of Michigan, Ann Arbor: Institute for Socia l Research.

PAGE 106

106 Monitoring the Future national survey results on drug use, 1975 2016: Overview, key findings on adolescent drug use. The University of Michigan, Ann Arbor: Institute for Social Research. Joreskog, K. G. and Sorbom, D., 1984. LISREL VI user's guide (3rd ed.). Mooresville, IN: Scientific Software. Juon, H. S., Fothergill, K. E., Green, K. M., Doherty, E. E., Ensminger, M. E., 2011. Antecedents and cons equences of marijuana use trajectories over the life course in an African American population. Drug Alcohol Depend 118 216 223. Kandel, D. B., 1996. The parental and peer contexts of adolescent deviance: An algebra of interpersonal influences. J. Drug I ssues 26 289 315. Kann, L., MCManus, T., Harris, W. A., Shanklin, S. L., Flint, K. H., Hawkins, J., Queen, B., Lowry, R., Olsen, E. O., Chyen, D., Whittle, L., Thornton, J., Lin. C., Yamakawa,Y., 2016. Youth risk behavior surveillance United States, 201 5. MMWR. Surveill Summ. 65(No. SS 6). Centers for Disease Control and Prevention (CDC), U.S. Department of Health and Human Services, Atlanta, GA. Keyes, K. M., Vo, T., Wall, M., Caetano, R., Suglia, S. F., Martins, S. S., Galea, S., Hasin, D., 2015. Raci al/ethnic differences in use of alcohol, tobacco, and marijuana: Is there a cross over from adolescence to adulthood? Soc. Sci. Med. 124, 132 141. Keyes, K. M., Wall, M., Feng, T., Cerda, M., Hasin, D. S., 2017. Race/ethnicity and marijuana use in the Uni ted States: Diminishing differences in the prevalence of use, 2006 2015. Drug Alcohol Depend. 179, 379 386. Kilm er, B., MacCoun, R., 2017. How medical marijuana smoothed the transition to marijuana l egalization in the United States. Annu. Rev. Law. Soc Sc i. 13 King, K. A., Vidourek, R. A., Hoffman, A. R., 2012. Sex and grade level differences in marijuana use among youth. J. Drug Educ 42 (3), 361 377. King, K. M., Fleming, C. B., Monahan, K. C., Catalano, R. F., 2011. Changes in self control problems an d attention problems during middle school predict alcohol, tobacco, and marijuana use during high school. Psychol. Addict Behav. 25 69 79. Korhonen, T., Huizink, A. C., Dick, D. M., Pulkkinen, L., Rose, R. J., Kaprio, J., 2008. Role of individual, peer and family factors in the use of cannabis and other illicit drugs: A longitudinal analysis among Finnish adolescent twins. Drug Alcohol Depend. 97 33 43.

PAGE 107

107 Kosterman, R., Hawkins, J. D., Guo, J., Catalano, R. F., Abbott, R. D., 2000. The dynamics of alcoh ol and marijuana initiation: Patterns and predictors of first use in adolescence. Am. J. Public Health 90 360 366. Kousha, M., Shahrivar, Z., Alaghband Rad, J., 2012. Su bstance use disorder and ADHD: I en Disord.16, 325 32. Lasopa, S. O., Striley, C. W., Cottler, L. B., 2015. Diversion of prescription stimulant drugs among 10 18 year olds. Curr. Opin. Psychiatry 28 292 298. Leung, R. K., Toumbourou, J. W., Hemphill, S. A., 2014. The effect of peer i nfluence and selection processes on adolescent alcohol use: a systematic review of longitudinal studies. Health. Psych Rev. 8, 426 457. Lisdahl, K. M., Wright, N. E., Medina Kirchner, C., Maple, K. E., Shollenbarger, S., 2014. Considering cannabis: The ef fects of regular cannabis use on neurocognition in adolescents and young adults. Curr. Addict. Rep. 1, 144 156. Loke, A. Y., Mak, Y., 2013. Family process and peer influences on substance use by adolescents. Int. J. Environ. Res. Public health. 10, 3868 3 885. with mental health symptoms and substance use in young adolescents. BMC Psychiatry 1 2 116. MacKay, A. P., Duran, C., 2007. Adolescent health in the United States, 2007. National Center for Health Statistics. Mandara, J., Rogers, S. Y., Zinbarg, R. E., 2011. The effects of family structure on J. Marriage Fam. 73 557 569. Marmorstein, N. R., White, H. R., Loeber, R., Stouthamer Loeber, M., 2010. Anxiety as a predictor of age at first use of substances and progression to substance use problems among boys. J. Abnorm. Child Psychol 38 211 224 Martins, S. S., Storr, C. L., Alexandre, P. K., Chi lcoat, H. D., 2008. Adolescent e cstasy and other drug use in the National Survey of Parents and Youth: the role of sensation 919 933 McCabe, S. E., Morales, M., Cranford, J. A., Delva, J., McPhe rson, M.D., Boyd, C. J., 2007. Race/Ethnicity and Gender Differences in Drug Use and Abuse Among College Students. J. Ethn. Subst Abuse. 6, 75 95.

PAGE 108

108 Miech, R., J ohnston, L.D., O'Malley, P. M., Bachman, J. G., Schulenberg, J., 2016. Monitoring the Future National Survey Results on Drug Use, 1975 2015: Volume I, Secondary School Students. The Uni versity of Michigan, Ann Arbor: Institute for Social Research. Miller, S. M., Siegel, J. T., Hohman, Z ., Crano, W. D., 2013. Factors mediating the Psychol. Addict. Behav. 27 848 853. Mongro Wilson, C., 2007. The Influence of parental warmth and con trol on Latino adolescent alcohol use. Hisp. J. Behav Sci 30, 89 105. Morrison, D. M., Lohr, M. J., Beadnell, B., Gillmore, M. R., Lewis, S., & Gilchrist, L., 2010. Planned Behav ior. Psychol Health 25 569 587. Moss, H. B., Chen, C. M., Yi, H., 2014. Early adolescent patterns of alcohol, cigarettes, and marijuana polysubstance use and young adult substance use outcomes in a nationally representative sample. Drug Alcohol Depend. 136, 51 62. Mrug, S., Gaines, J., Su, W., Windle, M., 2010. School Level Substance Use: Effects on J. Stud. Alcohol Drugs 71 488 495. Musick, K., Seltzer, J. A., Schwartz, C. R., 2008. Neighborhoo d norms and substance use among teens. Soc. Sci. Res. 37 138 155. NIDA, 2016. Substance use in women. Retrieved from https://www.drugabuse.gov/publications/research reports/substance use in women on 2017, September 5. Nelson, S. E., Ryzin, M. J., Dishi on, T. J., 2015. Alcohol, marijuana and tobacco use trajectories from age 12 to 24 years: Demographic correlates and young adult substance use problems. Dev Psychopathol. 27, 253 277. P., 2011. The association between substance use disorders and early and combined use of alcohol and marijuana in two American Indian populations. J. Subst Use 16 213 229. Ottosen C., Petersen, L., Larsen, J.T., Dalsgaard, S., 2016. Gender differences i n a ssociations between Attention Defi cit/Hyperactivity Disorder and substance u se. J. Am. Acad. Child Adolesc Psychiatry. 55, 227 234.

PAGE 109

109 Pacek, L. R., Malcolm, R. J., Martins, S. S., 2012. Race/ethnicity differences between alcohol, marijuana, and co occurr ing alcohol and marijuana use disorders and their association with public health and social problems using a national sample. Am. J. Addict 21 435 444. Pearson, M. R., Bravo, A. J., Conne r, B. T., 2017. Distinguishing subpopulations of marijuana u ser s with latent profile analysis Drug Alcohol Depend. 172, 1 8. Peiser, N. C., Heaven, P. C., 1996. Family influences on self reported delinquency among high school students. J. Adolesc. 19, 557 568. Perez, A., Ariza, C., Sanchez Martinez, F., Nebot, M., 2 010. Cannabis consumption initiation among adolescents: A longitudinal study. Addiction. 35, 129 134. Perkonigg, A., Goodwin, R. D., Fiedler, A., Behrendt, S., Beesdo, K., Lieb, R., Wittchen, H.U., 2008. The natural course of cannabis use, abuse and depen dence during the first decades of life. Addiction 103 439 449. Pinchevsky, G. M., Arria, A. M., Caldeira, K. M., Garnier Dykstra, L. M., Vincent, K. B., O Parent and p e er i nfluences. Prev. Sci 13 43 54. Ryan, S. M., Jorm, A. F., Lubman, D. I., 2010. Parenting factors associated with reduced adolescent alcohol use: A systematic review of longitudinal studies. Aust. N Z. J Psychiatry. 44, 774 783. Sagar, K. A., Dahlgren M. K., Gonenc, A., Racine, M. T., Dreman, M. W., Gruber, G. A. 2015. The impact of initiation: Early onset marijuana smokers demonstrate altered stroop performance and brain activation Dev Cogn Neurosci. 16, 84 92. Schepis, T. S., Desai, R. A., Cavallo D. A., Smith, A. E., McFetridge, A., Liss, T. B., Potenza, M. N., Krishnan Sarin, S., 2011. Gender differences in adolescent marijuana use and associated psychosocial characteristcs. J. Addict. Med 5, 65 73. Shacham, E., Cottler, L.B., 2010. Sexual beh aviors among club drug users: prevalence and reliability. Arch. Sex. Behav. 39, 1331 1341. Shih, R. A., Miles, J. N. V., Tucker, J. S., Zhou, A. J. thnic differences in adolescent substance use: Mediation by individual, fami ly, and school factors. J. Stud. Alcohol Drugs. 71, 640 651. Sibley, M. H., Pelham, W. E., Jr., Molina, B. S. G., Coxe, S., Kipp, H., Gnagy, E. M., Lahey, B. B., 2014. The role of early childhood ADHD and subsequent CD in the initiation and escalation of adolescent cigarette, alcohol, and marijuana use. J. Abnorm Psychol. 123, 362 374.

PAGE 110

110 Silins, E., Horwood, L. J., Patton, G. C., Fergusson, D. M., Olsson, C. A., Hutchinson, D. M., Mattick, R. P., 2014. Young adult sequelae of adolescent cannabis use: an i ntegrative analysis. Lancet Psychiatry 1 286 293. Steiger, J. H., 1990. Structural model evaluation and modification: an interval estimation approach. Multivariate Behav. Res. 25, 173e180. Stephens, P. C., Sloboda, Z., Stephens, R. C., Teasdale, B., G rey, S. F., Hawthorne, R. D., Williams, J., 2009. Universal school based substance abuse prevention programs: Modeling targeted mediators and outcomes for adolescent cigarette, alcohol and marijuana use. Drug Alcohol Depend 102 19 29. Striley, C. W., Ke lso Chichetto, N. E., Cottler, L. B., 2017. Nonmedical prescription stimulant use among girls 10 18 years of age: Associations with other risky behavior. J. Adoles. Health 60 328 332. Substance Abuse and Mental Health Services Administration, 2011. Resu lts from the 2010 National Survey on Drug Use and Health: Summary of National Findings (No. HHS Publication No. (SMA) 11 4658), NSDUH Series H 41. Substance Abuse and Mental Health Services Administration, Rockville, MD. Substance Abuse and Mental Health Services Administration, 2013. Drug Abuse Warning Network, 2011: National Estimates of Drug Related Emergency Department Visits. HHS Publication No. (SMA) 13 4760, DAWN Series D 39. Center for behavioral Health Statistics and Quality, Rockville, MD. Subst ance Abuse and Mental Health Services Administration, 2017. Results from the 2016 National Survey on Drug Use and Health: Summary of National Findings. Substance Abuse and Mental Health Services Administration, Rockville, MD. Svensson, R., 2000. Risk fact ors for different dimensions of adolescent drug use. J. Child Adoles. Subst Abuse 9 67 90. Swift, W., Coffey, C., Carlin, J. B., Degenhardt, L., Patton, G. C., 2008. Adolescent cannabis users at 24 years: trajectories to regular weekly use and dependenc e in young adulthood. Addiction 103 1361 1370. Terry among young adults: parallel trajectories in a national cohort sequential study. Addiction 106 1855 1865. Tu, A. W. Ratner, P. A., Johnson, J. L., 2008. Gender differences in the correlates of Subst. Use Misuse 43 1438 1463.

PAGE 111

111 van der Pol, P., Liebregts, N., de Graaf, R., Korf, D. J., van den Brink, W., van Laar, M., 2013. Predicting the t ransition from frequent cannabis use to cannabis dependence: a three year prospective study. Drug Alcohol Depend. 133 352 359. Vidourek, R. A., King, K. A., Montgomery, L. Psychosocial determinants of marijuana use among African American youth. J. Ethn. Subst Abuse. 16, 43 65. Villanti, A., Boulay, M., Juon, H., 2011. Peer, parent and media influences on adolescent smoking by developmental stage. Addict. Behav. 36, 133 136. Viveros, M.P., Marco, E.M., File, S.E., 2006. Nicotine and cannabinoids: Parall els, contrasts and interactions. Neurosci. Biobehav. Rev. 30, 1161 1181 Vogel, M., Rees, C.E., McCuddy, T. Carson, D.C., 2015. The highs that bind: School context, social status and marijuana Use. J. Youth Adolesc. 44, 1153 1164. Volkow, N. D., Baler, R. D., Compton, W. M., Weiss, S. R. B., 2014. Adverse health effects of marijuana use. N. Engl J. Med. 370 2219 2227. Walker, D. D., Neighbors, C., Rodriguez, L. M., Stephens, R. S., Roffman, R. A., 2011. Social norms and self efficacy among heavy using adolescent marijuana smokers. Psychol. Addict Behav. 25, 727 732. Wang, Y., Cottler, L. B., Striley, C. W., 2015. Differentiating patterns of prescription stimulant medical and nonmedical use among youth 10 18 years of age. Drug Alcohol Depend. 157 83 8 9. White, H. R., McMorris, B. J., Catalano, R. F., Fleming, C. B., Haggerty, K. P., Abb ott, R. D., 2006. Increases in alcohol and marijuana use during the transition out of high school into emerging adulthood: The effects of leaving home, going to c ol leg e, and high school protective f actors. J. Stud. Alcohol 67 810 822. Witbrodt, J., Mulia, N., Zemore, S. E., & Kerr, W. C., 2014. Racial/ethnic disparities in alcohol related problems: Differences by gender and level of heavy drinking. Alcohol. Cli. Exp. Re s. 38 1662 1670. Wright, N. E., Scerpella D, Lisdahl, K. M., 2016. Marijuana use is associated with behavioral approach and depressive symptoms in adolescents and emerging adults. PLoS One. 11, e0166005. Wu, L. T., Blazer, D. G., Swartz, M. S., Burc hett, B., Brady., K. T., Workgroup, N.A., 2013. Illicit and nonmedical drug use among Asian Americans Native Hawaiians/Pacific Islanders, and mixed race individuals Drug Alcohol Depend. 133, 360 367.

PAGE 112

112 Wu, L. T., Swartz, M. S., Brady, K. T., Hoyle, R. H., & Workgroup, N. A., 2015. Perceived cannabis use norms and cannabis use among adolescents in the United States. J. Psychiatr. Res. 64, 79 87. Wymbs, B. T., McCarty, C. A., King, K. M., McCauley, E., Vander Stoep, A., Baer, J. S., Waschbusch, D. A., 2012. Callous unemotional traits as unique prospective risk factors for substance use in early adolescent boys and girls. J. Abnorm Child Psychol 40 1099 1110. Yabiku, S. T., Marsiglia, F. F., Kulis, S., Parsai, M. B., Becerra, D., Colle, M. del., 2010. Par ental monitoring and changes in substance use among Latino/a and Non Latino/a pre adolescents in the Southwest. Subst Use Misuse 45 2524 2550. Zapolski, T. C., Fisher, S., Banks, D. E., Hensel, D. J., Barnes Najor, J., 2016. Examining the protective eff ect of ethnic identity on drug attitudes and use among a diverse youth population. J. Youth. Adolesc. 46, 1702 1715.

PAGE 113

113 BIOGRAPHICAL SKETCH S hivani R. Khan received her Bachelor of Science in biology from Salem College in Winston Salem, North Carolina in 2 010. After graduation, she worked as a Research Hospital in Boston, Massachusetts. She joined the Doctor of Philosophy (Ph.D.) program in epidemiology in the College of Me dicine and College of Public Health and Health Professions at the Unive rsity of Florida (UF) in August 2013. She was trained under the mentorship of Catherine W. Str iley, Ph.D., M.S.W., M.P.E., an associate professor, who directs the Master of Science in Epidemiology and the Certificate in Psychiatric Epidemiology programs. She received her Ph.D. in epidemiology from the University of Florida in the fall of 2017. Shiv among youth. Her research focus for dissertation was on predictors of daily marijuana use among US youth. As a Graduate Research Assistant in the Department of Epidemiology at UF, s he has gained extensive qualitative and quantitative research experience working on multiple health projects. She was a part of the team that developed widely used assessments, the Substance Abuse Module (SAM 5) and the Diagnostic Interview Schedule (DIS 5 ). She worked on the Post Earthquake Needs Assessment in Nepal, Haiti Health Study, ASSIST study, and Tics and Tourette Study. She also worked as a Student Assistant at the Department of Health Services Research, Management & Policy at UF. In that capacity she conducted data analyses using 2014 Florida Behavioral Risk Factor Surveillance System (FBRFSS) and 2015 People with Disability (PWD) survey and prepared the Disability Data Report describing demographic characteristics, health behaviors, health care utilization, chronic illness, and barriers to

PAGE 114

114 care among Floridian older adults with and without disabilities. She worked as a Teaching Assistant for three courses in the Department of Epidemiology at UF. Shivani has authored and co authored a number of pe er reviewed journal articles and a book chapter in the field of substance use and presented her research in internal as well as external conferences. In addition, she has received numerous awards including travel grants from (1) NIDA Women and Sex/Gender D ifferences Junior Investigator, (2) Center for Responsible Gaming (NCRG), (4) the Gr aduate Student Council at UF (5) Office of Research at UF, and (6 Further, s he received scholarships from Summer Institute in Infectious Diseases, University of Washington and University of Florida International Center.