PREVALENCE OF AND RISK FACTORS FOR NON MEDICAL USE OF PRESCRIPTION STIMULANTS : A NATIONAL STUDY OF YOUTH By SONAM ONGMU LASOPA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2015
2015 Sonam Ongmu Lasopa
4 ACKNOWLEDGMENTS As a two ye ar old, I moved or rather my father relocated my entire family to a neighboring state just so I could have a better education and continue to have the not what ships were meant for and that r oots and responsibilities are just as important if not more Since then, I moved on to several states and a country in pursuit of the best education and although my family did not move every place with me, they have been there in all possible ways -fanning my wings and giving me root edness I am what I am and all that I have achieved because of the gracious and unconditional support from my parents sisters brother s in law, nieces and nephew. I am truly blessed in this regard and want to acknowledge their wonderful impact in my life Along my academic journey, I have come to understand that the finest teacher s are not those who teach something, but a person who inspires the student to give the best so as to discover her potential while lead ing by example. My mentor Dr. Linda B. Cottler is one such teacher whom I look up to and for whom my admiration grows each day I have immense gratitude for her i nvest ment of time, effort and warmth in my life and development as a researcher I hope to carry on her legacy to improve access to care and services for the most underserved with passion and persistence I am as grateful to Dr. Catherine W. Striley for her part in my intellectual growth and progress She has been there for me in more ways than one -as a teacher who has so much to offer, as a caring friend who made my stay in the US easier and as in times of need. Dr. Almut Winterstein has also shaped the direction of my learning trajectory of which I am very appreciative Her astute insight and feedback in the present work has been very important to me as has been her interest in my future
5 goals I also extend my t hankfulness to Dr. Sarah Jo Nixon for her expert guidance in my work. I admire her unfathomable energy and effervescence which I wish to imbibe along with her deep understanding of the field of substance abuse research Overall, I hold in great value this group of inspiring teachers who have supported m e and provide d me with much needed guidance to the present work my learning and professional development I am honored to have been their student. And as we move towards an objective, it is as important to keep attention to the road because it is the road that teaches and enriches us as we walk in length. It is also equally essential to have fellow travelers who share the journey and add joie de vivre. I have had many such friends who became my family in Gainesville : Shivani R. Khan, Sunil Swami, P oulomy Chakraborty, Diana P. Rojas, Johanny Perez, Abenaa Ac heampong, Bimota Nambam Mirsada Ser dare vic and Vicki Osborne I cherish them dearly for the constant fun laughter and friendship I must thank everyone who w as part of the National Monitoring of Adolescent Prescription Stimulant s Study ( N MAPSS ) because of which this PhD dissertation is possible. Special gratitude is reserved for the 11 048 youth respondents for shar ing their exper iences and time i n the study. Their participation has enriched my work and learning to a vast extent. Lastly, I gratefully acknowledge the NIH Fogarty grants ( D # 43 TW09120 & D # 43 TW05811 ; PI Cottler ) that have supported my p re doctoral period at both the Washington University in St Louis and University of Florida This experience has brought me immense learning, research knowledge and growth over the course of the last five years I hope to justify this opportunity to the fullest extent possib le
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF FIGURES ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRA CT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 14 Prevalence ................................ ................................ ................................ .............. 14 Risk Factors of Non Medical Use of Prescription Stimulants ................................ .. 16 Sociodemographic Factors ................................ ................................ ............... 17 Externalizing and Internalizing Problem Behaviors ................................ .......... 18 Peer Influence ................................ ................................ ................................ .. 20 Age of Onset and Non medical Use of Prescription Stimulants .............................. 20 Non medical Use of Prescription Stimulants and the Use of Other Substances ..... 21 Limitations of Prior Non medical Prescription Stimulant Use Literature .................. 22 2 DATA SOURCE ................................ ................................ ................................ ...... 25 The National Monitoring of Adolescents Prescription Stimulants Study .................. 25 N MAPSS Assessment ................................ ................................ ........................... 26 Demographics ................................ ................................ ................................ .. 27 Mental and Behavioral Health Indicators ................................ .......................... 27 Peer Influence ................................ ................................ ................................ .. 28 Stimulant Use in the N MAPSS ................................ ................................ .............. 29 3 EXTERNALIZING AND INTERNALIZING RISK FACTORS OF NON MEDICAL USE OF PRESCRIPTION STIMULANTS: A NATIONAL STUDY OF YOUTH ....... 30 Background ................................ ................................ ................................ ............. 30 Externalizing Problem Behaviors and Non medical Use of Prescription Stimulants ................................ ................................ ................................ ..... 31 Internalizing Problem Behaviors and Non medical Use of Prescription Stimulants ................................ ................................ ................................ ..... 32 Gender and Non medical Use of Prescription Stimulants ................................ 33 Methods ................................ ................................ ................................ .................. 35 Measures ................................ ................................ ................................ .......... 35 Prescription stimulant use ................................ ................................ .......... 35 Externalizing problem behaviors ................................ ................................ 36
7 Attention deficit hyperactivity disorder (ADHD) ................................ .......... 36 Conduct disorder (CD) symptoms ................................ .............................. 36 Substance use ................................ ................................ ........................... 37 Internalizing problem behaviors ................................ ................................ 37 Depressive symptoms ................................ ................................ ................ 37 Anxiety symptoms ................................ ................................ ...................... 38 Weight issues ................................ ................................ ............................. 38 Peer influence ................................ ................................ ............................ 38 Demographics ................................ ................................ ............................ 38 Statistical Analyses ................................ ................................ .......................... 39 Results ................................ ................................ ................................ .................... 41 Prevalence of Externalizing and Internalizing Problem Behaviors and Non medical Use of Stimulants ................................ ................................ ............. 42 Association between Externalizing and Internalizing Behaviors and Non medical Use of Stimulants ................................ ................................ ............. 44 Discussion ................................ ................................ ................................ .............. 45 Prevalence of Exte rnalizing and Internalizing Behaviors and Non medical Use of Stimulants ................................ ................................ .......................... 46 Association of Externalizing and Internalizing Behavio rs and Non medical Use of Prescription Stimulants ................................ ................................ ...... 48 The Role of Gender in the Association between Externalizing Behaviors and No n medical Use Of Stimulants ................................ ............................. 51 4 AGE OF INITIATION AND RISK FACTORS OF NON MEDICAL USE OF PRESCRIPTION STIMULANTS ................................ ................................ ............. 58 Background ................................ ................................ ................................ ............. 58 Substance Use Initiation and Progression Theories ................................ ......... 58 Externalizing Problem Behaviors and Risk of Initiating the Non medical Use of Stimulants ................................ ................................ ................................ 60 Current Study ................................ ................................ ................................ ... 62 Methods ................................ ................................ ................................ .................. 63 Sample ................................ ................................ ................................ ............. 64 Inclusion criteria ................................ ................................ ......................... 64 Exclusion criteria ................................ ................................ ........................ 64 Measures ................................ ................................ ................................ .......... 65 Event measure ................................ ................................ ........................... 65 Age of non medical stimulant use initiation ................................ ................ 65 Length of follow up ................................ ................................ ..................... 66 ADHD ................................ ................................ ................................ ......... 66 CD symptoms ................................ ................................ ............................ 66 Other substance use initiation ................................ ................................ .... 67 Covariates ................................ ................................ ................................ .. 67 Statistical Analyses ................................ ................................ .......................... 68 Results ................................ ................................ ................................ .................... 69 Characteristics of Y outh with Non medical Use of Prescription Stimulants ...... 69 Hazard of Initiating Non medical Use of Stimulants ................................ ......... 70
8 Effect of Externalizing Problem Behaviors on the Initiation of Non medical Use of Prescription Stimulants ................................ ................................ ...... 71 Effect of Covariates on the Initiation of Non medical Use of Prescription Stimulants ................................ ................................ ................................ ..... 71 Sensitivity Analysis ................................ ................................ ........................... 72 Discussion ................................ ................................ ................................ .............. 72 5 NON MEDICAL USE OF PRESCRIPTION STIMULANTS AND THE USE OF OTHER SUBSTANCES AMONG YOUTH: A LATENT CLASS ANALYSIS ............ 82 Background ................................ ................................ ................................ ............. 82 Methods ................................ ................................ ................................ .................. 85 Sample and Procedures ................................ ................................ ................... 85 Measures ................................ ................................ ................................ .......... 86 Substance use ................................ ................................ ........................... 86 Socio demographic characteristics ................................ ............................ 88 Mental and behavioral health ................................ ................................ ..... 88 Peer substance use ................................ ................................ ................... 89 Analyses ................................ ................................ ................................ ........... 89 Results ................................ ................................ ................................ .................... 90 Substance Use Patterns Among Youth ................................ ............................ 90 LCA Defined Subtypes of Substance Use among Youth ................................ .. 91 Socio Demographic Characteristics by LCA Subtypes ................................ ..... 92 Mental and Behavioral Health Characteristics by LCA Subtypes ..................... 93 Peer Substance Use by LCA Subtypes ................................ ............................ 93 Multinomial Logistic Regression Analysis by LCA Subtypes ............................ 93 Discussion ................................ ................................ ................................ .............. 95 Non medical Use of Prescription Stimulants, Opioids, Benzodiazepines and Other Substance Use ( Rx++ ) ................................ ................................ ........ 96 Alcohol, Tobacco, Marijuana and Other Illicit Drug Users (ATM+) ................... 98 Alcohol and Marijuana Users ................................ ................................ ............ 98 Low no Substance Users ................................ ................................ ................. 99 Study Limitations and Strengths ................................ ................................ ..... 100 6 CONCLUSION ................................ ................................ ................................ ...... 106 APPENDIX A NATIONAL MONITORING OF PRESCRIPTION STIMULANT S SURVEY ........... 113 B SENSITIVITY ANALYSIS ................................ ................................ ..................... 136 LIST OF REFERENCES ................................ ................................ ............................. 137 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 156
9 LIST OF TABLES Table page 3 1 Socio demographic, mental and behavioral characteristics by stimulant use among yout h 10 to 18 years of age in the N MAPSS ................................ ......... 54 3 2 Socio demographic, mental and behavioral health problems by gender and stimulant use among youth 10 to 18 years of age in the N MAPSS ................... 55 3 3 Logistic regression models of externalizing and internalizing problem behaviors as risk factors for stimulant use among youth 10 to 18 years o f age in the N MAPSS (n=10984) ................................ ................................ ................ 56 4 1 Characteristics of youth aged 10 to 18 years by lifetime prescription stimulants use status. ................................ ................................ ......................... 78 4 2 Life time substance use among youth with non medical prescription stimulants use (n=467). ................................ ................................ ...................... 78 4 3 Life table estimates of hazard probabilities of initiation into non medical use of stimulants among youth (N=7691) aged 5 to 18 years from 2009 to 2011. .... 78 4 4 Estimated hazards of non medical prescription stimulant use initiation among youth 10 to 18 years of age in the N MAPPS ................................ ..................... 79 5 1 Substance use among youth in the N MAPSS (n=11,007). .............................. 102 5 2 Statistical fit measures for substance use subtypes of youth in the N MAPSS (n=11,007). ................................ ................................ ................................ ....... 102 5 3 Average latent class probabilities for most likely latent class membership for substance use (row) by latent class (column) youth in the N MAPSS. ............. 102 5 4 Demographic characteristics of substance use subtypes of youth in the N MAPSS. ................................ ................................ ................................ ............ 102 5 5 Health characteristics of substance use subtypes of youth 10 to 18 years of age in the N MAPSS. ................................ ................................ ....................... 103 5 6 Correlates of substance use subtypes of youth in the N MAPSS (n=11,007). .. 103 B 1 Estimated hazards of lifetime prescription stimulant use initiation among youth 10 to 18 years of age in the N MAPPS ................................ ................... 136
10 LIST OF FIGURES Figure page 2 1 Recruitment Sites of the N MAPSS ................................ ................................ .... 28 2 2 Past 30 day Stimulant Use among Youth in the N MAPSS ................................ 29 3 1 Conceptual diagram of the association between externalizing behaviors and the non medical use of prescription stimulants among youth. ............................ 57 3 2 Conceptual diagram of the moderation model with ADHD as an example ......... 57 4 1 Hazard estimates of age of initiation into non medical use of prescription stimulants among youth 10 to 18 years of age. ................................ .................. 80 4 2 Hazard estimates of age of initiation into non medical use of prescription stimulants by Attention Deficit Hyperactivity Disorder (ADHD) status among youth 10 to 18 years of age. ................................ ................................ ............... 80 4 3 Hazard estimates of age of initiation into non medical use of prescription stimulants by Conduct Dis order (CD) status among youth 10 to 18 years of age. ................................ ................................ ................................ .................... 81 4 4 Hazard estimates of age of initiation into non medical use of prescription stimulants by gender among youth 10 to 18 years of age. ................................ 81 5 1 Item probability for latent class subtypes of youth in the four class model (N=11,007) ................................ ................................ ................................ ....... 105
11 LIST OF ABBREVIATIONS A BIC Adjusted Bayes information criterion ADHD Attention deficit hyperactivity disorder AIC Aikake information criterion AOR Adjusted odds ratio AT M+ Alcohol, tobacco marijuana an d other drugs BLRT Bootstrapped likelihood ratio test CD Conduct disorder CI Confidence interval LCA Latent class analysis LMR LRT Lo Mendell Rubin adjusted likelihood ratio test MI Multiple imputation MTF Monitoring The Future MU Medical use N MAPSS National Monitoring of Adolescents Prescription Stimulant s Study NMU Non medical use NSDUH National Survey on Drug Use in Households OTC Over the counter SAM Substance Abuse Module VIF Variance inflation factor
12 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 PREVALENCE OF AND RISK FACTORS FOR NON MEDICAL USE OF PRESCRIPTION STIMULANTS : A NATIONAL STUDY OF YOUTH B y Sonam Ongmu Lasopa Dece mber 2 0 1 5 Chair: Linda B. Cottler Major: Epidemiology The non medical use (NMU) of prescription stimulants among youth is of public health concern due to its dependence potential and deleterious physical health consequences. Increased understanding of what risk factors influence the use of stimulants for non medical purposes among youth is of great need to adequately deal with the prescription drug epidemic in the United States. The present study aimed to examine the: 1) prevalence of and risk factors in terms of externalizing and internalizing problem behavio rs for past 30 day NMU of stimulants among youth 10 to 18 years; 2) period of highest hazard for initiation into non medical use of stimulants and factors that influence the onset and 3) empirically determined subtypes of substance use among youth. Data comes from the National Monitoring of Prescription Stimulant Study (N MAPSS) which recruited 11,048 youth between 10 to 18 years of age using an entertainment venue intercept method. The cross sectional study was carried out in ten cities across the United States from 2008 to 2011. Assessments included lifetime and past 30 day use of prescription stimulants, the quantity, routes and reasons for stimulants use, use of cigarettes alcohol, marijuana and other illicit substance use
13 mental health and behavioral factors, peer use of stimulant s and demographic information The main outcome measure in the study is the non medical use of prescription stimulants defined as use of stimulant medication without a prescription, in greater amou nts, more often, or longer than prescribed, or for a reason other than a doctor said one should take them. Youth with a history of NMU of stimulants w ere compared to youth with a history of medical use only and non users. The study is an in depth investig ation into the risk factors of non medical use of prescription stimulants with a focus on gender specific characteristics. Findings will indicate when youth are at highest risk of initiating stimulant use non medically and inform the field of the sub type s of youth most likely to misuse. The work will have important implications in the design of targeted prevention and intervention programs to reduce non medical use of prescription stimulants among youth in the United States.
14 CHAPTER 1 INTRODUCTION Prescription stimulant medications such as methylphenidate, dextroamphetamine and amphetamine salts are used in the treatment of Attention Deficit Hyperactiv ity Disorder (ADHD), narcolepsy, obesity and occasionally depression (Bogle and Smith, 2009; Challman and Lipsky, 2000; Visser et al., 2007) These medications are listed as Schedule II d rugs because of their potential for non medical use and dependence. Non medical use (NMU) refers recommended dose, route or use without a prescription (Cottler et al., 2013) Following NMU, tolerance for stimulants can d evelop along with the risk of addiction Additionally, the use of stimulants for non medical purposes has been shown to cause several medical issues such as cardiovascular problems (rapid or irregular heartbeat, increase in blood pressure, stroke), increas ed body temperature, convulsions and even death (Lakhan and Kirchgessner, 2012) Prevalence The past two decades in the United States ha ve been marked with a significant growth in the rates of prescription stimulant medications paralleled by increased rates of non medical use (Garfield et al., 2012; Johansen et al., 2015; Visser et al., 2007) Young adults and adolescents have been consistently shown to be at highest risk for NMU ; a pproximately a third of the 2.4 million new initiat ors of NMU of prescription drugs including stimulants within the past year were between the ages of 12 and 17 years (Substance Abuse and Mental Health Services Administration, 2011; Teter et al., 2005) Several national and regional studies have reported the pervasiveness of NMU leading to its des ignati on as a public health problem of grave concern among youth.
15 Overall, among youth between 10 to 18 years of age the prevalence of past year non medical use of stimulants ranges from 5 10% with past 30 day use up to 5% (Cott ler et al., 2013; Johnston et al., 2012; McCabe et al., 2007; McCabe et al., 2004; Nakawaki and Crano, 2012; Schepis and Krishnan Sarin, 2008; Simoni Wastila et al., 2008a; Substance Abuse and Mental Health Services Administr ation, 2014; Viana et al., 2012 ) For instance, Monitoring the Future (2012) which is a national student based stud y indicate d a growing trend in the NMU of stimulants among 12 th grade students; past year rates of amphetamine abuse increased from 6.8% to 8.7% over the last five years from 2008 to 2013. A slight decline was observed for those in the 8 th (0.3%) and 10 th grade (0.6%). During the same period, an increase in the rates of current NMU for 12th grade students was also observed (from 2.9 to 4.1%) while rates remained stable for 8 th and 10 th graders at around 1.4% and 2.8% respectively. Using combined data from a n annual cross sectional household survey for the years 2002 2009 Sweeney and others (2013) found lifetime NM U of stimulants to be 3.4% for those aged 12 years and older. Other smaller or regional samples report higher rates NMU of stimulants compared to those in national samples (McCabe et al., 2015, 2012; McCabe and West, 2013) For instance, a study found the lifetime rate of all non medical use of prescription drugs to be 6.5% among 6 th to 12 th graders in public schools at Mississippi. Stimulants (e.g., Ritalin, Concerta, Focalin, Dexedrine; 37%), were the third most common drug of misuse after pain medications (57%) and benzodiazepines (44%) (Viana et al., 2012a) Also, indications of diversion -t he selling and sh aring of prescription drugs to individuals without prescriptions have been reported; with friends or family being most commonly
16 reported as the main source for their NMU (Clemow and Walker, 2014; Cottler et al., 2013; Fischer et al., 2010; Ka ye and Darke, 2012) For example, one in seven adolescent students reported having received or exchanged prescribed medications from their friends for non medical purposes (West et al., 2011) It is clear that there is significant supply of and access to ADHD medication available to youth in the US indicating potential for NMU and its adverse effects (Setlik et al., 2009; Valdez, 2014) A cause of added concern is the significant rise in emergency department (ED) visits related to stimulant non medical use. In the years 2004 to 2009, ED visits increased by 102% for methylphenidate and by 276% for amphetamines (Substance Abuse and Mental Health Services Administration, 2011) Further more the Drug Abuse and Warning Network (DAWN) surveillance system developed to monitor morbidity and mortality as sociated with drug use indicates that stimulant related ED visit rates have steadily risen over both the long term (307%) and short term (85%) while rates for other medications such as opioids and benzodiazepines have remained stable (Substance Abuse and Mental Health Services Administration, 2013) Because of the risk of NMU and its deleterious co nsequences, there have been attempts to control the availability and accessibility of prescription drugs with measures (such as prescription monitoring systems) which relate to limiting the supply of prescription medication s for NMU (Fischer et al., 2010; Zosel et al., 2013) What is less known is the demand aspect of NMU among youth the identification of youth at risk for non medical use of stimulants Risk F actors of Non Medical Use of Prescription Stimulants Developmentally, youth differ from adults in maturation and behavior reflected in part as differential patterns and correlates of substance use (Cotto et al., 2010) While the majority of NMU of stimulants literature has focused on the college going population,
17 prior studies do suggest that developmental life stage specific risks exist and risk factors vary depending on the population and setting which necessitates the characterization of youth with NMU (Clark et al., 2013; DeLi si et al., 2015; Moss et al., 2014; Piehler et al., 2012 ; Herman Stahl et al., 2008) S ome important risk factors shown i mportan t in prior substance use literature have been disproportionately understudied in NMU of stimulants focused studies (such as mental health) while a few others have been well researched (such as sociodemographic variables other substance use perceiv ed risk or harm from use ) (McC abe and West, 2013; Nargiso et al., 2015; Sweeney et al., 2013; Teter et al., 2005) A brief review of risk factors is presented in the following sections. Sociodemographic F actors The m ajority of previous studies on NMU of stimulants among youth have f ocused on sociodemographic risk factors such as gender, race, family characteristics and academic performance (Chen et al., 2014; Collins et al., 2011; King and Chassin, 2007; McCabe et al ., 2012; McCabe and West, 2013; Nakawaki and Crano, 2012) Specifically, being of White race has been consistently found to be associated with higher rates of NMU of stimulants in youth while those of other race s such as African American s and Asians are at lower risk (Conn and Marks, 2014; McCabe et al., 2004; McCabe and West, 2013) With regard to the role of gender, the evidence so far is largely inconclusive; some studies have found increased rat es among females, others report higher rates in males while o thers indicate non differential risk by gender (F leary et al., 2011; McCabe and West, 2013; Nakawaki and Crano, 2012; Sweeney et al., 2013) There is considerable evidence to suggest that NMU of stimulants among youth is associated with negative consequences such as school dropout, low academic perfor mance and low educational attainment (Arria et al., 2011; Collins et al., 2011;
18 Garnier Dykstra et al., 2012; McCabe et al., 2012; Nargiso et al., 2015; Schepis and Krishnan Sarin, 2008; Young et al., 2012a) Additionally, parental cha r acteristics are also important; i n particular, poor monitoring and involvement by parents is positively related to NMU of stimulants in youth in the majority of the studies (Ford and McCutcheon, 2012; Schinke et al., 2008; Vaughn et al., 2012) Also, youth from dual parent households have been consistently found to be at decrea sed risk for NMU of stimulants compared to youth from a single parent or absent parent households (Kaye and Darke, 2012; Young et al., 2012 ) E xternalizing a nd Internalizing Problem Behaviors To better understand childhood mental health and behavior and its implications, a large body of research distinguishes between externalizing and internalizing behavior problems (Achenbach and Edelbrock, 1978) Externalizing behavior problems denote nal environment and include key problems of disruptive behavior, hyperactivity, aggression and delinquency (Campbell et al., 2000; Eisenberg et al., 2001) Attention Deficit H yperact ivity Disorder (ADHD), Conduct D isorder (CD) and Oppositional Defiant Disorder (ODD) fall in the scope of externalizing behavior problems. In contrast, internalizing behavior problems affect the internal psychological environment of an individual and are characterized by behaviors such as social withdrawal, negativity, anxiety, inhibition and depression (Achenbach and Edelbrock, 1978; McCulloch et al., 2000) Internalizing disorders encompass the diagnoses of Anxiety D isorder and Major Depressive Disorder (Campbell et al., 2000; Eisenberg et al., 2001; Hinshaw, 1987) Although the dichotomy of externalizing and internalizing behaviors ha s been proposed, there is evidence that these behav iors can occur concomitantly (Kirisci et al., 2015;
19 Miettunen et al., 2014; Scalco et al., 2014; Weiland et al., 2014) Mor e importantly, both externalizing and internalizing behavior problems have been significantly associated with substance abuse. For example, long term associations exist between childhood externalizing behavior problems and abuse of substance such as alcoho l, tobacco and other illegal drugs (Helstrom et al., 2004; King et al., 2004) Internalizing behavior problems in childhood have been shown to increase the risk of initiating use of illegal substances in young adolescen ts (King et al., 2004) The associatio n of externalizing and internalizing behavior problems and NMU of stimulants has not yet been fully examined; a lack of nationally representative data that investigate s mental health correlates of NMU of stimulants among youth has been suggested as an area of need (Sung et al., 2005) In terms of externalizing problem behaviors, a few studies have found that antisocial behavior and delinquent activities are linked to increased rates o f NMU of stimulants (Chen et al., 2014; Gilson and Kreis, 2009; Harrell and Broman, 2009; McCauley et al., 2010; Nargiso et al., 2015; Vaughn et al., 2012; Viana et al., 2012) Among the internalizing problems only depression has been studied as a correlate of NMU of stimulants in youth. A national study by Goldstein (2008) found that among youth between the ages of 12 and 17 with NMU of stimulants at least 20% reported having experienced an episode of major depression in the past 12 months. A recent study also reports a positive association between the NMU of prescription medication s and major depressive disorder s ( MDD ) ; an excess risk of up to 35% for MDD was found among users relative to non using youth (Ali et al., 2015) Most of the
20 other internalizing problems (such as anxiety) have not been subjected to investigation in th e context of NMU of stimulants in youth Peer I nfluence A number of studies have demonstrated a significant relationship between peer or friend approval of substance use and NMU of medication among youth (Ford et al., 2014; Ford and Lacerenza, 2011) Peer disapproval of substance use seems to be a protective factor that decreases the likelihood of NMU of stimulants while youth with friends that have increasingly tolera nt attitudes towards the use of substances seem to be at greater risk of having engaged in non medical use of stimulant s in the past 12 months (Ford and Lace renza, 2011; Young et al., 2012 ) Moreover, the use of substances by a close friend has been consistently seen as an important risk factor for NMU of stimulants in line with most other substance use research. For example, Schinke and colleagues (2008) found females w ho had best friends who used substance s were over five times at increased risk for NMU of stimulants than their counter parts. Age of O nset and N on m edical Use of Prescription Stimulants Adolescence is a critical developmental stage for the structural and functional maturation of the human brain; this period is marked by significant changes in cognition and behavior such as risk taking, and the emergence of substance use and psychological pathologies. The m ajority of the substance use literature has shown that the most vulnerable years for initiating substance use are between the ages of 12 and 20 (Substance Abuse an d Mental Health Services Administration, 2011) Detrimental effects of substance use are greater on the developing adolescent brain than they are among adults. Moreover, among those who use, the risk of developing substance
21 dependence is two to three ti mes higher when onset is in adolescence t han when youth begin use later (Brook et al., 2007; Gil et al., 2004; King and Ch assin, 2007) While there are several studies that have examined the onset and consequences of substance use such as alcohol, tobacco, marijuana, few have focused on prescription stimulants (Moss et al., 2014; Seedall and Anthony, 2013) An e arlier age of initiation of substance use is linked to a progression into l ong term substance use patterns, especially stimulants. Previous studies on the onset o f NMU of stimulants have focused on reporting the mean age of first stimulant use as an important factor However, i nformation regarding the critical period of risk for onset of NMU of stimulants based on general population youth samples are lacking ; this knowledge would be of importance for preventi on or for messages that focus on delay ing initiation of stimulant non medical use While longitudinal studies are more appropriate to examine predictive relations, cost and logistic considerations make suc h studies difficult and unfeasible to conduct. Non medical Use of Prescription Stimulant s and the Use of Other Substance s Youth who engage in NMU of stimulants a lso report increased use of other drugs compared to those who use stimulants medically or abst ain. Several studies have revealed associations between NMU of stimulants and increased rates of use of cigarettes; heavy episodic drinking; m arijuana, and other illicit drug use among youth in the United States (Arria et al., 2011; Collins et al., 2011; Garnier Dykstra et al., 2012; McCabe et al., 2012; Nargiso et al., 2015; Schepis and Krishnan Sarin, 2008; Young et al., 2012 ) There are indications that characteristics of youth who engage pr imarily in NMU of stimulants significantly differ from youth who engage in comorbid or poly substance use A handful of studies have empirically classif ied youth into distinct groups
22 determined by the ir patterns of substance use or by motives and routes of use (McCabe et al., 2009; McCabe an d Cranford, 2012) Findings from young adult and college going age populations and those focused on other drug classes besides stimulants suggest that there may be important but unrecognized differences between substance use subtypes ; the possible varia tion in risk correlates for different groups of youth stimulant users is still unknown (Hall et al., 2010; McCabe et al., 2009 ; Lamont et al., 2014; Moss et al., 2014; Roth et al., 2015) Knowledge of N MU of stimulants related taxonomies among youth and their associated characteristics can contextualize NMU of stimulants within the larger public health scope of substance use and abuse behaviors Limitations of Prior Non medical Prescription Stimulant Use Literature A lack of representativeness limit s the existing studies to date in our und erstand ing of the association of risk factors and NMU of stimulants The m ajority of studies have focused on students and restricted geographical locations or are web based (Boyd et al., 2009, 2007; Viana et al ., 2012 ; Webb et al., 2013) Two national studies -the National Study on Drug Use and Health (NSDUH) and the Monitoring the Future (MTF) hav e assessed NMU among youth; however, variable definitions of NMU inclusion of only a few stimulant medications and varying methodology make comparisons with other studies difficult. The MTF although large is school based and does not include youth who are school drop outs or students who missed school the day of the survey, this may be a significant proportion of the very youth who are at increased risk for substance use (Bracken et al., 2013 ; Cottler et al., 2013) Additionally, these large samples are not suited to examine characteristics of NMU of stimulants among youth because they are not detailed in their assessment of factors that can influence the non medical use of stimulants.
23 Because of these reasons such as the variable definitions of non medical use and lack of generalizability of study findings, there is a gap in the understanding of what factors influence prescription stimulant use among youth NMU of stimulants continues to be a problem among youth, with about 5% 10% reporting past 12 month of stimulant misuse (McCabe et al., 2014; We yandt et al., 2014 ; Wang et al., 2015 ) The present study will provide a comprehensive and in depth investigation into the non medical use of stimulants among youth across ten US cities in order to narrow the knowledge gap on risk factors for NMU of st imulants One unique aspect of the study is the sample; data come s from the National Monitoring of Adolescent Prescription Stimulant Study (N MAPSS) which i s the only study to focus primarily on stimulant use patterns among youth 10 18 years of age. In fac t, youth as young as 10 to 12 years of age were assessed for patterns of stimulant use and the field currently lacks data for this age group Further, the present study uses a sample that is large assesses use of the majority of commonly prescribed stimul ants among respondents recruited from across the contiguous US a long with a number of important potential correlates based on extant substance use literature Therefore, the study is ideally suited to add to the current prescription stimulant misuse litera ture among youth. Additionally, the emphasis of the present study is placed on a holistic understanding of NMU of stimulants and focuses on the period from when youth are at highest risk to the most current past 30 day non medical use and the identification of factors that place youth at increased risk for NMU of stimulants An important component is the attention on the role of gender in the NMU of stimulants in youth. The empirical identification of subtypes of youth based on thei r substance use patterns is
24 also a highlight of this work because it aims to contextualize the NMU of stimulants with in the larger concept of substance use and misuse The proposed work will be important in the comprehensive and overall understanding of no n medical use of stimulants among youth which h as value in the design of strategic prevention and intervention programs The ultimate aim is to utilize the study findings t o reduce substance abuse among youth.
25 CHAPTER 2 DATA SOURCE The National Monitoring of Adolescents Prescription Stimulants Study Data for the present study comes from the National Monitoring of Adolescent Prescription Stimulants Study (N MAPSS) which surveyed youth 10 to 18 years of age. An entertainment venue intercept method ology was used to recruit youth across ten US cities. T he s tate with the highest volume of stimulant prescriptions within one of ten federal regions (OMB Circular A105) was first selected; within each state the city with highest volume (shown in Figure 2 1 ) was then selected from which recruitment was carried out (Boston, New York, Philadelphia, Tampa, St Louis, Cincinnati, Houston, Denver, Los Angeles and Seattle). Eligibility criteria included: being 10 18 years of age residing in an urban, suburban or rural ZIP code from one of the ten cities. Youth unaware of their ZIP code, non English readers, cognitively impaired, or college going were excluded. Recruiter interviewers (RI) visited entertainment venues such as malls, parks and other recreational plac es that are youth friendly identified potential respondents (based on the study eligib ility criteria) and introduced the study. Consent for participation in the N MAPSS was verbally obtained B e cause data were anonymous parental permission/informed consen t was deemed to be not required by the Institutional Review Board ( IRB ) at Washington University where the study commenced Recruiter interviewers approached 21,444 youth for study participation. Among those approached, 25% did not stop to hear about the s tudy, an additional 21% were not eligible and 10% refused to participate. Among those who participated (11,468), 420 were excluded because the surveys were duplicates, incomplete, ineligible or questionable with regard to quality or veracity of responses. The final sample consisted
26 of 11,048 youth that represented an 86.7% participation rate and a 68% overall response rate T he N MAPSS survey assessed for recognition and patterns of use of prescription stimulant medication along with other substance use ( ci garettes alcohol, marijuana and other illegal drugs). Data collection was carried out over four cross sections between Fall 2008 and Spring 201 1 Recruitment goals sought at adequate representation of all ages of interest as well as urban, suburban and rural areas. Demographic characteristics of the N MAPSS sample were comparable to those of the US 2010 census across the same age range (Cottler et al., 2013) Participants were each remunerated with a gift card worth $10 f rom a national electronics store. The study approval was obtained from the IRB at Washington University in St Louis for all four cross sections of the study and University of Florida I RB for the fourth cross section. All analyses for the present study used data from all cross sections except for one component that examined age of onset of stimulant use that was assessed only in the latter three cross sections. N MAPSS Assessment The N MAPPS survey was adapted from the 1) Substance Abuse Module (SAM) (Hort on et al, 2000) found to be reliable for q uantity and frequency of stimulant and other substance use and 2) the Washington University Risk Behavior Assessment (RBA) for risk factors for non medical use (Shacham and Cottler, 2010) The assessment consisted of two parts. Part I of the survey assessed for demographics along with mental and behavioral variables (described in detail below). Then, youth were presented with picture s of 5 dosages of commonly prescribed amphetamines, 2 pictures of other common stimulant me dication and a common over the counter (OTC) pain medication and a decongestant. Youth were asked if they had ever seen the
27 medication; if they answered positively, they were asked to identify the stimulant by brand name and formulation. Part II started with a presentation of pictures of five stimulants -Adderall or Adderall XR Concerta Ritalin Daytrana and Vyvanse individually, by dose and formulation. Sp ecific questions asked for: a) lifetime and past 30 day use of each stimulant; b) use of stimulant s more than prescribed; c) use that belonged to someone else (parents, brother or sister, different family member, someone from school or from work, someone u nknown and other); d) route of ingestion of stimulant (by mouth, snorted or sniffed, smoked and other); e) reasons for prescription stimulant use and f) age of first use of each stimulant contingent on whether youth responded to using the stimulant. Questi ons also assessed if each stimulant endorsed for use had been prescribed by a psychiatrist or a doctor. Demographics The N MAPSS assessed for age, gender, race and ethnicity. Zip code level information was used to determine eligibility and to categorize area of residence as urban, suburban or rural. The survey also included items that elicited whether youth lived in a dual parent household grades in school, sleep timings during the week, number of meals in a week youth ate with family and self rated gene ral level of health (excellent, good, fair and poor). Mental a nd Behavioral Health Indicators A number of mental and behavioral health pr oblems were measured. 1) self reported ADHD diagnosis; 2 ) conduct disorder symptoms : g etting into a lot of trouble at home or school, r unning away from home overnight, being arrested, us ing or threaten ing someone with a weapon suspension from school for any reason ; 3 ) w eight
28 issues : ever been very afraid of gaining weight, ever tried to lose weight by making yourself vom it, taking pills, not eating for a day or two and exercising too much; 4 ) d epressi ve symptoms : past 12 month loss of interest for at least 2 weeks, and feeling sad or depressed in the past 12 months for at least 2 weeks; 5 ) anxiety: lifetime worry or stres s lasting 6 months or more ; 6) s ubstance u se : i ndividual questions for both lifetime and past 30 day use of cigarettes, alcohol, marijuana and lifetime use of other illegal drugs that included cocaine, heroin, ecstasy, LSD, steroids, inhalants and methamphetamine. Further, youth were asked if they had ever used the prescription medication Xanax Vicodin Vyvanse or Oxycodone individually. Additionally, age of first use of each of the substances was assessed contingent on whether youth reported having used the substance. Peer I nfluence In order to a sses for peer influence on stimulant use, youth were asked for the number of close friends who had used the stimulant Adderall at least once. F igure 2 1 Recruitment Sites of the N MAPSS
29 Stimulant Use in the N MAPSS For the purposes of this study, past 30 day stimulant use was categorized into a three level variable : NMU of stimulants, medical use (MU) of stimulants and no use. P ast 30 day NMU of stimulants was defined as the use of any of the five stimulants b y non oral routes (except Daytrana), use that belonged to someone else, use of any The m edical use (MU) of stimulants was defined as the use of any of the five stimulants with a prescri ption within the prescribed dosage, use only by oral routes (except Daytrana which comes in a patch) and no use of stimulant that belonged to someone else. Medical use was cross tabulated with NMU and those youth who endorsed both (118) were considered NM U Youth who did not report past 30 day use of any s timulants were categorized as non users of prescription stimulant medications. Using these definitions, 6.8% of youth reported some form of stimulant use in the past 30 days ; as shown in Figure 2 2, 3.6% (398) reported NMU and 3.2% (334) were MU. Figure 2 2. Past 30 day Stimulant Use among Youth in the N MAPSS 93.2% 3.2% 3.6% No Use Medical Use Non-medical Use
30 CHAPTER 3 EXTERNALIZING AND INTERNALIZING RISK FACTORS OF NON MEDICAL USE OF PRESCRIPTION STIMULANTS : A NATIONAL STUDY OF YOUTH Backgroun d Prescription stimulant medication such as methylphenidate, dextroamphetamine and amphetamine salts have been highly effective in the treatment of Attention Deficit Hyperactivity Disorder (ADHD), however, the abuse potential of these medications and repor ts of misuse of prescription stimulants by youth in the United States has generated much public heal th concern in the past decade (Chen et al., 2014; Clemow and Walker, 2014; Johnston et al., 2014, 2012; McCabe et al., 2011; Miech et al., 2015; Schepis and Krishnan Sarin, 2008; Simoni Wastila et al., 2008; Substance Abuse and Mental Health Services Administration, 2014; Sweeney et al., 2013; Viana et al., 2012 ; Young et al., 2012) Approximately 5 10% of youth aged 10 to 18 years in the US engage in the misuse or non medical use of prescription stimulant medication s (NMU) defined as the use of medications by non oral routes, use of medication that belongs to someone else, overuse of own medication, or use motivated by getting high or experimentation of the stimulant medication (Cottler et al., 2013; Lasopa et al., 2015 ; Wang et al., 2015 ) Because the patterns and correla tes of substance use abuse and dependence in youth differ from the adult population, it is important to understand developmental specific risks of NMU (Cotto et al., 2010) While some efforts to deepen our understanding of what risk factors are linked to NMU of stimulants among youth have been conducted, the effor ts have disproportionately focused on s ocio demographic factors, illicit drug use, perceived risk or harm from stimulant use; other individual level factors such as mental health characteristics that have been shown to be important in other substance use a mong youth have been understudied (McCabe and West, 2013;
31 Nakawaki and Crano, 2012; Nargiso et al. 2015; Sweeney et al., 2013) It has been suggested that distinct factors place some youth at increased likelihood for non medical use and these factors var y by medication p opulation and setting (Herman Stah l et al., 2007) While m ental health problems have play ed an important role in comorbidity of other substance use ( tobacco, alcohol and other drugs ) they have not been the focus of research among youth in the context of stimulant use and misuse (Barkley et al., 2004; Brook et al., 2007; Flory et al., 2003; Gil et al., 2004; Marmorstein, 2009; Marmorstein et al., 2010) In conceptualizing child and adolescent mental health, a distinction has been made between two types of problem behaviors termed as externalizing and interna lizing behavior s (Achenbach, 1966) Externalizing problems are behaviors manifest overtly and associated with problems in controll ing unwanted or disruptive behaviors, aggressive behaviors and hyperactivity (Liu, 2004) These are disorders of Attention Deficit Hyper activity Disorder (ADHD), Oppositional Defiant Disorder (ODD) and Conduct Disorder (CD). In contrast, internalizing problems are related to the internal psychological processes characterized by an inability to effectivel y regulate emotions and they include social withdrawal, somatic complaints, sadness, anxiety and depression (Liu, 2003) In general, externalizing problems find increased representation among males while the prevalence of internalizing problems is high among females. Although substance use has been conceptualized as a form of externalizing behavior, there is evidence to suggest it is r elat ed to internalizing problems as well. Externalizing P roblem Behaviors a nd Non m edical U se of Prescription Stimulants Externalizing problems such as C D problems and antisocial traits among youth have been linked to NMU; prevalence rates are elevated among those with antisocial and delinquent activities (Gilson and Kreis, 2009; Harrell and Broman, 2009; McCauley
32 et al., 2010; Nargiso et al., 2015; Vaughn et al., 2012; Viana et al., 2012a) For instance, Viana and others (2012 ) reported that youth were 1.5 times more likely to have engaged in NMU of prescription medication if they had been involved in physical fights. S everal studies indicate the role of ADHD in contributing to the risk of substance use across varied samples and methodological diffe rences (Charach et al., 2011; Lee, 2011; Molina et al., 2007; Riggs, 2014) A recent study by Harstad and Levy (2014) showed that children with ADHD were two times mor e likely to develop substance use disorders as compared to those without ADHD. Additionally, ADHD is known to pose increased risks for academic, social and psychological difficulties among youth relative to their non ADHD peers Further, students diagnosed with ADHD have indicated having to read material over and over to understand. Although a large majority of youth report NMU of stimulants to improve their academic or cognitive performance, youth are as likely to endorse using stimulants to get high (Boyd et al., 2006) Internalizing Problem Behaviors and Non m edical Use o f Prescription Stimulant s Internalizing problems such as anxiety and depression have been implicated in the developme nt of NMU of prescription medication s For example, more than one in five youth aged 12 to 17 years in a national study who met criteria for NMU of stimulants had experienced an episode of major depression in the past 12 months (Goldstein, 2008) Further, a known side effect of prescription stimulants is weight loss and there are reports that a proportion of youth misuse these medications to control weight often cit ing weight loss as a motive. Youth have also indicated the use stimulants to deal with negative emotion and stress related eating behaviors (Jeffers and Koester, 2013; Terry McElrath et al., 2009)
33 Overall, the literature suggests that youth with externalizing ( attention difficulties, hyperactive behaviors, conduct or antisocial traits ) or internal izing problems ( emotional problems ) may be at increased risk of NM U of stimulant s. Gender and N on medical U se of P rescription S timulants Patterns of substance use abuse and dependence are known to vary by gender and several studies have examined gender distribution in the prevalence of NMU of stimulants The m ajority of analyses from NSDUH report i ncreased rate s of NMU of stimulant s among females compared to males in the age range of 12 to 17 years (Herman Stahl et al., 2008; Nakawaki and Crano, 2012) Other studies have found greater NMU of stimulant s among males, still others indicate no difference by gender suggesting that the role of gender in NMU of stimulant s is still not clear (Chen et al., 2014 ; King et al., 2013; McCabe et al., 2014; Sweeney et al., 2013; Teter et al., 2005) Females have been shown to be more likely than males to abuse substances when they are concerned about peer approval (Rohrbach and Milam, 2002) Although some efforts to understand risk factors for NMU of stimulants among youth in general are underway the intersection between mental health and the NMU of stimulant s, and the role of gender in this relationship still needs to be addressed A concep tual model illustrating the relationship between externalizing and internalizing behaviors with the NMU of stimulants among youth along with the possible moderating role of gender i s presented in Figure 3 1. The m ajority of prior research on mental health and NMU of stimulant s among youth ha ve not included the general population or out of treatment populations; studies that do focus on general populations have assessed non medical use of stimulants but only within a broader category of non medical use of prescription medications In order
34 to address important ga ps in the literature for NMU of stimulant s literature and to extend previous work, the present study aims to examin e a broad range of mental and behavioral risk factors of NMU of stimulants in terms of externalizing and internalizing problem indicators with a specific focus on the role of gender with a large national sample of youth. First, we aim to: 1) assess the prevalence and association of externalizing ( ADHD and CD symptoms a nd past 30 day use of c igarette s alcohol, marijuana and lifetime other illegal drug use) and internalizing problems ( depressive and anxiety symptoms, and weight concerns ) by stimulant s use status The hypotheses for this aim are: 1 a) ADHD symptoms will in crease the risk of NMU of stimulants among youth; 1b) youth with CD symptoms will have increased risk of NMU of stimulants; 1c) past 30 day use of c igarette s alcohol, marijuana and lifetime other illegal drug use will be positively associated with NMU of stimulants in youth ; 1d ) depressive symptoms will increase the risk of NMU of stimulants among youth; 1e) youth with anxiety symptoms will have increased risk of NMU of stimulants and 1f) we ight concerns will be positively associated with NMU of stimulants among youth Additionally the dissertation focuses on two externalizing problem behaviors and aims to : 2 ) ex amine the role of gender in the associations between ADHD and CD, and the NMU of stimulants in youth Fo r this aim we hypothesize that 1) male gender will increase the association between ADHD and NMU of stimulants than when youth are female ; 2) male gender will increase the association between C D and NMU of stimulants CD symptoms than when youth are female
35 Methods Data came from a cross sectional survey known as the National Monitoring of Prescription Stimulants Study (N MAPSS). The N MAPPS was con ducted in the years 2008 to 2011 and recruited a sample of 11,048 youth aged 10 to 18 years from entertainment venues in ten cities across the United States. The sample was found to be highly representative of the US population (Cottler et al., 2013) Interviewers surveyed youth anonymously on patterns of prescription stimulant and other substance use along with a number of socio demographic information. Measures Prescription s timulant u se The N MAPSS survey assessed use of prescription stimulants -Adderall or Adderall XR Concerta Ritalin Daytrana and Vyvanse individually. Pictures of each stimulant medication were presented and youth were asked to identify the stimulant by brand name and formulation. Specific questions asked for: a) lifetime and past 30 day use of each stimulant; b) use of stimulant s more tha n prescribed; c) use that belonged to someone else (parents, brother or sister, different family member, someone from school or from work, someone unknown and other); d) route of ingestion of stimulant (by mouth, snorted or sniffed, smoked and other), and e) reasons for prescription stimulant use. Questions also assessed if each stimulant endorsed for use had been prescribed by a psychiatrist or a doctor. For th is study p ast 30 day NMU of stimulants was defined as the use of any of the five aforementioned stimulants for use other than by mouth (except Daytrana ), use of any stimulant that belonged to someone else use of any stimulant more than prescribed or use
36 Medical use (MU) wa s defi ned as the use of any of the five stimulants with a prescription within the prescribed dosage, use only by oral routes (except Daytrana which comes in a patch ) with no use of any prescription stimulant that belonged to someone else. Medical use was cross tabulated with NMU and those youth who endorsed both (118) were consistently considered NMU as has been done in prior analyses ( McCabe and West, 2013; Cottler et al., 2013 ) Youth who did not report any of the five stimulants assessed wer e categorized as n on users of prescription stimulant medication s Using these definitions, the final sample includes a three level past 30 day stimulant use variable with 398 youth reporting any non medical use (NMU) 334 youth reporting use of prescription stimulants for m edical purposes only (MU) and 10, 252 youth reporting no use ( non user ) Ext ernalizing problem b ehaviors For the purposes of this study, symptoms of ADHD and CD, and substance use were considered externalizing problem behaviors. Attention deficit hyperactivity disorder (ADHD) ADHD was Has a doctor ever told you or your parents that you have Attention Deficit Disorder (ADD) or ADHD? Youth responded positively were considered to have ADHD considered no ADHD Conduct disorder (CD) symptoms Five items measured CD symptoms ever used or threatened
37 first three items were categorical and coded either yes or no ; the last two variables were coded yes when youth res ponded having at least one episode of having received tickets or warnings, or being arrested at least once They were coded no when there were none. Based on these items, a variable to represent conduct problems was created when youth positively responded to a minimum of three of the items. Substance use Questions assessed for both lifetime and past 30 day use of use of cigarette, alcohol, marijuana and other illegal drugs (cocaine, heroin, ecstasy, LSD, steroids, inhalants or methamphetamine). Dichotomou s variables for the past 30 days were each individual substance (cigarette, alcohol, marijuana) except for other illegal drug use which was only measured as a lifetime question Internalizing problem b ehaviors For the purposes of this study, anxiety and depression symptoms and weight issues were considered internalizing problem behaviors. Depressive symptoms D epressive symptoms were elicited with two items that asked : n the past 12 months have you had two weeks or more Respon ses to either of the two questions were considered as having depressi ve symptoms and coded yes; not responding positively was coded no
38 Anxiety symptom s A positive respon se was considered as having anxiety symptom s and coded 1 while a negative response was taken to indi cate having no anxiety and was coded 0 Weight i ssues Weight issues were measured with five questions : a) Making yourself vomit? b) Taking pills? c) Not eatin responses to any three of the five questions were considered to indicate having weight issues and coded 1 while two positive responses or less were coded as 0. Peer influence Peer stimulant use influen close friends have tried Adderall were coded as yes and those who reported having no friends w ere coded as no Demographics N MAPSS assessed for a number of demographic characteristics that include age, gender, race and Zip code of residence Zip code level information was used to categorize area of residence into urban, suburban and rural. The survey also included items that elicited characteristics within the home such as whether youth lived with both parents, sleep timings during the week (youth were dichotomized into those who regularly went to bed at 12 am or later and those who went to bed before 12 am) and number of times in a week youth ate me als with family (binary variable created with youth eating with family ate at least 5 times in a week coded as yes and those not
39 coded no ). Grades in school were also ass vs. D F Statistical A nalyses First, prevalence estimates for NMU of stimulants MU only of stimulants and no use by gender were computed. Chi square analyses were conducted to examine differences in sociodemographic, stimulant use status externalizing and internalizing disorders and peer influence by gender. Stratified chi square analyses were carried out to explore the effects of gender on past 30 day prescription stimulant use (NMU of stimulants MU only of stimulants and no use) by sociodemographic characteristics externalizing an d internalizing disorders and peer stimulant use influence. Next, bivariate logistic regression models were conducted to calculate unadjusted odds ratios between each individual risk factor and covariates and the outcome variable of NMU of stimulants desc ribed earlier. Then, multivariate logistic regression models were conducted to explore factors associated with increased risk of non medical prescription stimulant use with non users as the referent group for both NMU and MU only groups. A standard approach to model building was applied (Hosmer et al., 2013) ; only those variables in univariate models which were statistically associated (at p 0.05) with the outcome variable were retained in the model building process. In the multivariate model, variables were entered in the following order: socio demographics, mental and behavioral health factors, other substance use measures peer stimulant us e and interaction terms of gender with ADHD, and gender with C D The a ssessment of multi collinearity was c onducted to identify highly associated independent variables before their insertion into the models; this was conducted with a
40 regression model with all variables to estimate the t olerance and Variance Inflation F a ctor (VIF). Generally a tolerance below .40 is suggestive of multi collinearity ; VIF coefficient above 10 is indicative of a variable that is correlated with other variable s in the model (Davis et al., 1986) High tolerance and low VIF was observed indicating no evidence of multi collinearity of variables in the model. Following the inclusion of each variable, the decision to either drop or retain a variable wa s based on whether its coefficient differed significantly from 0 (adjusting for the effects of the other variables), whether removal of the variable altered the remaining coefficients of other terms in the model by more than 20% Also, it was consider ed wh ether change in the overall fit of the model was improved by its addition. In some cases, a variable was retained in a model for statistical significance but later became insignificant when other variables were dropped. In these few cases, the variable tha t became insignificant was retained in the model. Further, t wo way interaction terms were created to test the hypothesis of the moderating role of gender on the relationship between ADHD and CD variables and NMU in the model. In general, a moderator is a variable (in this case gender) that affects the direction and strength of the relationship between a predictor X and an outcome Y; the causal relationship between X and Y changes as a function of the moderator (Baron and Kenny, 1986) In th e s e anal yse s, the expected moderation model is shown in Figure 3 1 where g ender is a moderator (M) and two externalizing behavior s (ADHD, CD) are the predictor s when NMU is the outcome. Based on a conceptual model (Hayes, 2013) t he moderation model for the present study was conceptualized to test whether f emales were more likely to experience NMU of
41 stimulants compare d to males when they are experiencing the same level of externalizing symptoms (ADHD and CD) A diagrammatic representation of the expected mechanism involved in the relationship between the predictor X ( ADHD as an illustration) on outcome Y (NMU of stimulants ) and influence d by the moderator M (gender) is shown in Figure 3 2 It is displayed by the arrow from the moderator M to the line linking predictor X to outcome Y Only those interaction terms that were statistically associated (at p 0.05) with the outcome variable indicating moderating effects were retained in the final moderation model. A less conservative stepdown Bonferroni correction was used to account for multiple testing. About 5% of the sample had missing data for one or more externalizing or internalizing behaviors peer stimulant use variable s and the covariates of grades, sleep timings and household structure. Multiple imputation procedure s w ere used to impute values for the missing scores using the proc mianalyse command in SAS 9.4. The final model was assessed for goodness of fit in terms of the AIC, deviance scores and lackfit test. The final model showed lower AIC values compared to prior models ; deviance scores were non significant at 0.05 level of significance and lackf it values were high indicating adequate model fit. Adjusted odds ratios with 95% confidence intervals are reported and all analyses were carried out using SAS 9.4. Results Shown in Table 3 1 are characteristics of youth in the total sample and by stimulant use status. Compared to MU only and youth with no use of stimulants, those who reported NMU of stimulants were more likely to be male, Caucasian, had poorer live in ru ral areas and less likely to live in a dual parent household.
42 Prevalence of Externalizing and Internalizing Problem Behaviors and Non m edical Use of Stimulants Al s o shown in Table 3 1 are the rates of externalizing problems in youth by stimulant use statu s using a three level variable Overall, 3 .2 % of youth reported only medical use of stimulants in the past 30 days p rior to the N MAPSS survey; 3.6 % reported non medica l use of stimulants, while 93.2 % reported no use of any of the five stimulants assessed About 13% of youth reported having been diagnosed with ADHD; a similar rate of CD symptoms was reported by youth (12.7%). While ADHD was highest among those with MU only ( 83.8 %) 41% of youth with NMU of stimul ants also reported ADHD. A s ignificantly higher rate of CD was observed among youth with NMU of stimulants ( 42.8 %) compared to both MU only (18.9%) and no use (11.4%) Past 30 day use of cigarette s alcoho l or marijuana ranged from 9.9 % to 27.0 % and lifetime other illegal drug use w as reported by 11.7 % of youth Substance use among youth with NMU of stimulants was two to three times higher relative to MU only and up to six times higher than youth with no use. In terms of th e prevalence of internalizing behaviors, high rates of depressive symptoms (21.8% ) were observed About one fourth of youth reported anxiety symptoms (25.8%) and w eight concerns were reported by 3.8%. The rates of all three internalizing behaviors wer e higher among youth with NMU of stimulants as compared to MU only while those with no use had the lowest rates. About one third of youth reported having at least one friend with stimulant use and the rate was significantly increased among youth with NMU of stimulants at 83% and more than half with MU only reported having friends who used stimulants.
43 Further, we examined whether the socio demographic characteristics and prevalence of externalizing and internalizing behaviors among youth differed by gender with in each stimulant use category. As shown in Table 3 2, there was no gender difference in race or area of residence amon g youth with NMU of stimulants or MU only; however, a higher representation of Caucasian race was seen among females with no stimulant use compared to males. Males were more likely to report poor grades s) compared to females in all three levels of stimulant use In the NMU of stimulants group and those with no use, a higher proportion of males reported going to bed at 12 am or later most days of the week compared to fe males w hile no gender difference was observed in youth with MU only. We found that there were significantly higher rates of externalizing conduct problems among males compared to females among all stimulant use groups. The prevalence of ADHD was higher among males compared to fe males among youth with NMU of stimulants and among those with no use while there was no gender difference in the ADHD rate for youth with MU only Further, more females were likely to report cigarettes use relative to males in the NMU of stimulants group ; alcohol, ma rijuana and other illegal drug use was comparable across gender Among the no use group, males were more likely to report marijuana and other illegal drug use compared to females. No gender difference was observed in substance use in the MU only group. Wi th regard to internalizing behaviors, among youth with NMU of stimulants and no use b oth depressive and anxiety symptoms and weight concerns were overrepresented in females relative to males. It is seen that m ore than half of females with NMU of stimulants report depressive (53.6%) and anxiety (56.8%) symptoms while
44 about 30% females reported weight concerns; these are considerably high rates of internalizing behaviors Weight concerns were higher among females compared to males in the MU only and no use gr oups. Gender difference in p eer stimulant use rates w as observed only among the no use group where slightly more males reported having a friend who used stimulants (26% vs. 23.6%). Although, no gender based difference in peer stimulant use were seen among youth with NMU of stimulants, the majority of males and females with NMU of stimulants have at least one friend who used stimulants ( 84.7 and 81.1% respectively). A ssociation between Externalizing and Internalizing Behaviors and Non medical Use of Stimula nts As shown in Table 3 3 controlling for all other variables in the model, the strongest risk factor for non medical use of prescription stimulants w as use of stimulant which increased the odds of NMU of stimulants by more than five time s ( AOR 5 6 8, CI 4. 12 7 68 ) compar ed to not having a close friend who used stimulants This was followed by the externalizing behaviors of ADHD that increased the risk of NMU of stimulants by over three times compared to not having ADHD (AOR 3 17 CI 2 45 4 09 ) Further, lifetime use of other illegal drug s increased the likelihood of NMU of stimulants by almost three times (AOR 2. 94 CI 2. 23 3.8 8 ) compared to no use of other illegal drugs The only internalizing problem that increased the risk of NMU of stim ulants was having weight concerns compared to no weight concerns ( AOR 2 28 ; CI 1. 57 3 30). Additionally, past 30 day use of alcohol (AOR 1. 70 ; CI 1.27 2.26), cigarettes (AOR 1. 5 9; CI 1.09 2.32) and marijuana (AOR 1. 81 ; CI 1. 37 2. 40 ) were also positively as sociated with increased odds of NMU of stimulants relative to their counterparts. We did not find any gender based association between the externalizing behaviors of ADHD or CD and
45 the NMU of stimulants ; interaction terms were not statistically significant using the Bonferroni correction for multiple tests. Among the socio demographic characteristics assessed, t he likelihood of NMU of stimulants increased with Caucasian race (AOR 1.65) compared to non C aucasians while b eing male increased the odds of NMU of stimulants by 1.39 times relative to females. When results for MU only were examined with no use serving as the referent group ADHD symptoms had the strongest association with MU only with an AOR of 50.79 (CI 36.93 69.84) relati ve to th ose without ADHD which was as expected P eer stimulant use (AOR 2 35 ; CI 1 78 3 11 ) emerged as another important correlate of MU only among youth Further being of Caucasian race and living in suburban or rural areas compared to urban areas were positively associated with MU only while o lder age was negatively associated E ach year increase in age was associated with about 1 7 % decrease in MU only. Discussion In this nationally representative sample of over 11,000 youth, analyses resulted in some important findings. First we found high rates of externalizing behaviors of ADHD and CD symptoms, substance use a s well as the internalizing behaviors of depressive and anxiety symptoms, and weight concerns among non medi cal use rs of stimulants Overall rates were largely comparable to those reported by prior studies. Second, consistent with our hypothesis t he externalizing behaviors of ADHD and use of the substances cigarettes, alcohol, marijuana and o ther illegal drug s and only internalizing behavior s of weight concern had a significant main effect on the n on medical use of stimulants. However, the strongest association was between peer
46 stimulant use and non medical use of stimulants Third, we did not find evidence t o support the moderating role of gender in the relationship between the externalizing behaviors of ADHD and CD symptoms an d the non medical use of stimulants Prevalence of Externalizing a nd Internalizing Behaviors and N on m edical U se of Stimulants We found rates of self reported ADHD in our sample ( 12 5 vs. 10%) to be comparable to the reported ADHD rates in the general population (Castle et al., 2007; Merikangas et al., 2010; Vande Voort et al., 2014; Visser et al., 2007) However, youth rates of self reported conduct disorder symptoms (12.7 vs. 7%) were high compared to youth in other community studies (Merikangas et a l., 2010 ) On the other hand r ates of CD up to 13.5 % among boys and 5.4 % among girls have been reported (Bloom et al., 2013) One reason our study found high rates of CD symptoms may be explained by the manner in which CD symptoms were measure d. First, the measure s we used w ere self reported and second, it is possible that our measures most likely capture d subclinical and subjective symptoms of C D compared to other studies that are clinical or diagnostic Further, we found high rates of ADHD symptoms among youth with medical use only (83.8%) as well as non medical use ( 40 9 %) Although it is expected that medical users would report having been diagnosed with ADHD we are concern ed with a fairly high proportion of non medical use rs who also reported ADHD who were either overusing their medications or were using medications without prescriptions It is known that less than half of children in the community with ADHD receive mental he alth care over the past one year (Merikan gas et al., 2010 ) A national study showed that even when most youth with ADHD received some form of treatment (behavioral or medication
47 or both) the quality of care was not in compliance with best practice s for ADHD treatment (Visser et al., 2014) Poor access or inadeq uate care of those with ADHD direct s us to the self medication hypothesis which suggests th at youth may be us ing stimulants non medical ly to self treat their ADHD symptoms (such as in attention) that they experienc e as distressing. Prior work has demonstrated that youth mis use stimulants in their quest for aca demic and cognitive improvement and it is possible that academic achievement motives are related in part to the problematic symptoms of ADHD (Boyd et al., 2006; Rabiner et al., 2009) While the self medication of undiagnosed ADHD symptoms has not been commonly cited as a primary motiv e for non medical use of stimulants there have been reports of stimulant misuse among those who perceive that they had ADHD (Boyd et al., 2009) This explanation may also be linked to our finding of higher rates of CD symptoms among youth with non medical use of stimulants An estimated 60% of those with CD are likely to have at least one ment al health disorder or academic and learning problem (Bernstein, 2015) Despite the lack of clear causal etiology academic problems have been associated with conduct problems which may in turn influence the non medical use of prescription stimulants. In terms of substance use our past 30 day marijuana rate of 17.2% and alcohol use of 27% co rresponds to those reported in national studies ( up to 21% for marijuana and up to 37% for alcohol ) (Miech et al., 2015) However, youth in our study reported higher rates of cigarette use (10 vs. 3%) and lower rates of other illegal drug use ( 11.7 vs. 27.4 %) than other studies Other substance use (such as alcohol, marijuana, other prescription medication misuse and illicit drugs) has been commonly rep orted among
48 non medical users of stimulants. For instance, McCabe and West (2013) reported that the majority of studen ts with non medical use of prescription stimulants also had increased rates of non medical use of opioids, sedatives, or tranqu iliz ers One reason for the lower rates of other illegal drug use could be related to our sample which included youth at younger ages while youth in other samples were slightly older. For internalizing problems, our lifetime rate s correspond to those of ot her studies. Rates of anxiety ( 25.8% ) and depressive symptoms (21.8%) and weight concerns (3.8%) were comparable at around 31.9%, 30% and 3% respectively (Centers for Disease Control and Prevention 2014; Merikangas et al., 2010 ) However, rates of internalizing problems were high among those with non medical use of stimulants. Association of Externalizing and Internalizing Behaviors and N on medical U se of P rescription S timulants While y outh with externalizing behaviors of ADHD and use of cigarettes, alcohol, marijuana and other illegal drugs were more likely to use stimulants non medically ; ADHD and il legal drug use were the more important factors that increased the risk of non medical use of stimulants by about three times Se veral investigations have demonstrated positive relationship s between ADHD symptoms and non medical use of stimulants as well as other substance use (Arria et al., 2012; Bright, 2008; Cassidy et al., 2013; Dussault and Weyandt, 2013; Kollins, 2008; McCabe et al., 2007; Poulin, 2007) Further evidence from N SDUH (Substance Abuse and Mental Health Services Administration, 2008) suggests that t hose who use stimulants non medical ly have high er rates of alcohol or drug use disorders, increased use of illicit drugs or greater participation in mental health treatment compared to non users. I ncreased poly substance use (cigarettes, alcohol, marijuana ) among youth with non medical use of
49 stimulants has been shown across the general and student population, and among residential care settings (Chen et al., 2014; Hall et al., 2010; Rhoades et al., 2014) Although a lar ge proportion of youth in our study reported CD symptoms, when controlling for other substance use (other illegal drugs, cigarettes, alcohol, marijuana), the association of CD symptoms with non medical stimulant use did not hold true. While several studies suggest that conduct problems may concomitantly occur with substance use, substance use may also precede conduct problems which may be a more important factor for the non medical use of stimulants in youth For instance, one study indicated that the use o f tobacco, marijuana and cocaine prior to conduct problems such as theft, drug dealing or arrest ( Morihisa et al, 2007 ). However, the association of conduct problems, other substance use and non medical use of stimulant s in youth may require further exploration. It is noted that youth who only used stimulants medically only had similar rates of substance use as the non users; this suggests that youth in the general population who appropriately use their prescribed medication s for ADHD are not at highe r risk for substance misuse than non users of stimulants. Although the reasons for the link between ADHD and substance use and nonmedical use of prescription stimulants is not clear, a common factor model suggest s that the relationship is the result of sh ared risk factors (Hawkins, 2009) Both ADHD and substance use have been characterized as disorders of impaired control that are influenced by s imilar neurobiological deficits that are associated with impulsivity and sensation seeking (Arcos Burgos et al., 2012; Duka, 2011) Neuroimaging studies i ndicate abnormalities in the activation of anterior cingulate and fronto subcortical systems along with dopaminergic and striatal involvement (Casey and Jones, 2010;
50 Frodl, 2010) Although it is beyond the scope of this paper, future studies may examine this suggested association of ADHD and non medical use of stimulants among youth. It is of note that risk of non medical stimulant use increase d by five fold for youth with a friend with stimulant use com pared to those without a friend stimulant use This finding is consistent with a substantial body of research indicating t hat peer use and approval or disapproval of use, have a profound influence on the non medical use of stimulants and other substance ab use (Collins et al., 2011; McCabe and West, 2013; Schinke et al., 2008) Sung and colleagues (2005) ha ve shown that even the mere perception of drug use by peers is linked to actual drug abuse by youth. Further King and others (2013) found that both peer stimulant non medical use and use of other substances such as alcohol or marijuana a lso had similar positive effect s on non medical use. In terms of internalizing problems, we found that only weight concerns elevate d the risk of stimulants non medical use Stimulant s are known to have the side effects of appetite suppression and consequent weight loss and youth may be motivated to use stimulants non medically in order to lose weight (Kent et al., 1995; Zachor et al., 2006 ; Jeffers and Benotsch, 2014; Jeffers and Koester, 2013; Teter et al., 2005) One of the few studies that examined motives for prescription dru gs among students found about 4.5% of youth indicate d amphetamine use to reduce weight. Further, the risk was stronger among regular non medical users of stimulants (AOR 3.3) compared with occasional users (AOR 1.6) Females were also significantly more likely to report weight reduction motives than males (Terry McElrath et al., 2009)
51 Although a large proportion of youth in our study reported symptoms of depression or anxiety, when controlling for other fact ors the association with non medical stimulant use did not hold true Similar to our findings, the non medical use of prescription medication was not related to depressive symptoms among adolescents in grades 7 to 12 (Harrell and Broman, 2009) H owever, i n other studies of youth, major depressive disorder (MDD) has been linked to the non medical use of prescription drugs O ne example is the NSDUH study wherein, almost one in four youth who used stimulants in the past 12 months had a major depressi ve episode in the past year compared with less than 10% of youth who did not use stimulants (Goldstein, 2008) The available measure for depression and anxiety in our study was based on only one or two items which may not adequately capture the emotional constructs and this could be related to the discrep ant findings. The Role o f Gender i n the Association between Externalizing Behaviors and Non m edical Use Of Stimulants R esults of earlier studies particularly those using NSDUH data have found increased risk of non medical use of stimulants among females Thus we were motivated to examine whether gender specific mental health risk factors for non medical use of stimulants would emerge in a representative national sample of youth. Unlike some prior studies we found that boys aged 10 to 18 years of age w ere more likely to report the non medical use of prescription stimulant s compared to females (Nakawaki and Crano, 2012) However, there are also s tudies that similarly reported increased rate s of non medical use of stimulant s among males (Fleary et al., 2011; King et al., 2013; McCabe et al., 2014)
52 In this study, we did not find evidence to support our hypotheses that gender plays a moderating role in the associations between ADHD and non medical use of stimulants and CD a nd the non medical use of stimulants However, t wo prior studies have s how n gr eater a ssociation between tobacco use and risk of non medical use of prescription drugs (including stimulants) for females compared with males (Back et al., 2010; Berenson a nd Rahman, 2011) One possible reason we did not find differential risk in the association of ADHD or CD and non medical use of prescription stimulants by gender could be related to our measure of non medical use which only examined the rates but not th e patterns of non medical use by gender The study findings may be viewed in light of some limitations. First, our measures are based on self reports there may be concerns related to social desirability issues and errors in recall using self reported measures However, self reports of substance use have been shown to be reliable and used by a large number of studies to assess the non medical use of prescription drugs (Harrington and Newman, 2007) Our study also used pictures of the medications assessed to aid in brand recognition of the st imulants which may have minimize d However, w e used a screener for the diagnosis of ADHD, CD, and depression and anxiety symptoms and this may have contributed to higher or lower rates compared to those of other studies Moreover we only assessed anxiety symptoms in the last six months. Another limitation is related to the cross sectional nature of the study which limits our ability to make temporal associations Our study also has some important strengths W e had a large sample of over 11,000 that were representative of youth in the US population (Cottler et al. 2013)
53 Compared to other national studies, an entertainment recruitment venue method enabled the study sample to include non in residence student population s (such as sc hool drop outs and home schooled ) as well. Further, t he study also assessed for a broad range of mental and behavioral risk factors which has increased our overall understanding of non medical use of stimulants in youth. Despite the limitations noted the study identified a number of important risk factors of non medical use of stimulants. Knowledge of risk factors is relevant for prevention, intervention and public health aspects of stimulant use and misuse Programs to prevent non medical use of stimu lants from developing into stimulant dependence must be developed and the outcomes evaluated.
54 Table 3 1. Socio demographic, mental and behavioral characteristics by stimulant use among youth 10 to 18 years of age in the N MAPS S Characteristic No use (n=10252) MU only (n=334) NMU (n=398) p value Total Socio demographics Age (Mean) 15.1 14.7 16.2 <.0001*** 15.1 Male gender 47.3% 52.4 % 54.0 % 0.0077** 47.7% Caucasian 41.6% 60.5% 63.9% <.0001*** 43.0% Residence Urban 94.5% 2.2% 3.3% 47.6% Suburban 92. 8% 3.6% 3.7% <.0001*** 37.3% Rural 91.1% 4.4% 4.5% 15.1% Grades in school (Ds & Fs) 23.1% 30.2% 40.9% <.0001*** 24.0% 50.7% 52.2% 31.1% <.0001*** 50.1% Sleep timings later than 12 am 21.9% 22.7% 39.2% <.0001*** 22.6% Dual parent household 56.2% 49.4% 46.2% <.0001*** 55.7% Externalizing behaviors CD symptoms 11.4% 18.9% 42.8% <.0001*** 12.7% ADHD 9 0 % 83 8 % 40 9 % <.0001*** 12 5 % Cigarette use past 30 days 8.5% 15.1% 43.6% < .0001*** 9.9% Alcohol use past 30 days 25.3% 28.7% 69.7% <.0001*** 27.0% Marijuana use past 30 days 15.5% 20.4% 60.9% <.0001*** 17.2% Other illegal drugs lifetime use 9.7% 17.7% 58.3% <.0001*** 11.7% Internalizing behaviors Depressive symptoms 20.7% 30.6% 43.0% <.0001*** 21.8% Anxiety symptoms 24.8% 33.2% 45.7% <.0001*** 25.8% Weight concerns 3.3% 5.4% 16.6% <.0001*** 3.8% Peer influence Stimulant use by friend 24.7% 53.9% 83.0% <.0001*** 27.7% p <.05, ** p <.001, *** p <.0001 Note: ADHD Attention Deficit Hyperactivity Disorder; CD Conduct Disorder
55 Table 3 2 Socio demographic, m ental and behavioral health problems by gender and stimulant use among youth 10 to 18 years of age in the N MAP S S Characteristic No use (n=10252) MU only (n=334) NMU (n=398) Female Male p value Female Male p value Female Male p value Socio demographics Caucasian 44.0% 38.8% <.0001*** 59.5% 61.5% 0.7095 65.4% 62.6% 0.5677 Residence Urban 47.5% 48.9% 34.6% 34.3% 38.8% 47.4% Suburban 37.5% 36.6% 0.3988 43.4% 44.0% 0.9937 40.4% 35.3% 0.2185 Rural 15.0% 14.5% 22.0% 21.7% 20.8% 17.2% Grades in school (Ds & Fs) 17 9 % 28 .2% <.0001*** 20.7 % 38.9 % 0.0003** 35 .5% 45 6 % 0.0419* 5 per week) 49.7% 51.9% 0.0234* 50.9% 53.4% 0.6476 29.7% 32.2% 0.5814 Sleep timings later than 12 am 18.0% 26.2% <.0001*** 20.7% 24.6% 0.4060 32.8% 44.6% 0.0157* Dual parent household 57.0% 55.3% 0.0867 50.3% 48.6% 0.7503 44.8% 47.4% 0.5995 Externalizing behaviors CD symptoms 7.11% 16.1% <.0001*** 13.2% 24.1% 0.0110* 32.4% 51.6 % 0.0001*** ADHD diagnosis 6.9 % 11.5 % <.0001*** 83.0 % 84.6 % 0. 7003 33 8 % 47. 0 % 0.0 081 Cigarette use past 30 days 8.1% 8.8% 0.1952 15.2% 14.9% 0.9498 49.7% 38.4 % 0.0241* Alcohol use past 30 days 25.7% 24.9% 0.3649 32.5% 25.3% 0.1484 71.1% 68.6 % 0.5862 Marijuana use past 30 days 12.5% 18.7% <.0001*** 18.9% 21.8% 0.5017 57.8% 63.5 % 0.2473 Other illegal drugs lifetime use 8.3% 11.3% <.0001*** 15.7% 19.5% 0.3622 54.4% 61.6 % 0.1520 Internalizing behaviors Depressive symptoms 24.5% 16.6% <.0001*** 33.5% 28.0% 0.2731 53.6% 33.9 % <.0001*** Anxiety symptoms 29.1% 20.0% <.0001*** 32.1% 34.3% 0.6685 56.8% 36.3 % <.0001*** Weight concerns 5.2% 1.07% < .0001*** 10.1% 1.1% 0.0003** 29.5% 5. 6 % <.0001*** Peer influence Stimulant use by friend 23.6% 26.0% 0.0046* 51.3% 56.3% 0.3560 81.1% 84. 7 % 0.3485 p <.05, ** p <.001, *** p <.0001 Note: ADHD Attention Deficit Hyperactivity Disorder; CD Conduct Disorde r
56 Table 3 3. Logistic regression model s of externalizing and internalizing problem behaviors as risk factors for stimulant use among youth 10 to 18 years of age in the N MAP S S (n=10984) Characteristic No use Medical use only AOR (95% CI) Non medical use AOR (95% CI) Socio demographics Mean age (SD) ref 0.8 3 (0. 78 0. 8 9) 0.98 ( 0 91 1. 05 ) Male gender ref 0.99 ( 0 74 1.31 ) 1. 39 (1. 0 3 1 .9 0 ) Caucasian ref 1. 26 ( 0 97 1.64 ) 1. 53 (1.2 0 1 96 ) Residence ref Suburban vs. urban ref 1. 39 (1. 05 1 84 )* 0.9 0 (0.70 1. 17 ) Rural vs. urban ref 1. 47 (1. 03 2. 08 )* 1.0 0 (0.7 3 1. 39 ) ref 0.78 (0. 59 1.00) 1.0 1 (0. 78 1. 31 ) Meals with family (<=5 per week) ref 0. 8 8 (0. 67 1. 15 ) 1.16 (0.8 9 1. 49 ) Sleep timings later than 12 am ref 0. 86 (0. 63 1. 17 ) 0.9 9 (0.7 7 1.2 8 ) Lives in a d ual parent household ref 1. 0 8 (1. 84 1. 40 ) 0.9 5 (0.7 4 1.2 0 ) Externalizing behaviors CD symptoms ref 0 80 (0. 56 1. 14 ) 1. 26 ( 0 95 1. 67 ) ADHD diagnosis ref 5 0 79 (3 6 93 69 .8 4 ) 3 17 ( 2.45 4 09 ) Cigarette use past 30 days ref 1. 18 (0.6 7 2. 07 ) 1. 5 9 (1. 09 2. 32 ) Alcohol use past 30 days ref 0. 94 (0. 68 1. 30 ) 1.70 (1.2 7 2. 26 ) Marijuana use past 30 days ref 0.85 (0. 59 1. 2 3) 1. 81 (1. 37 2.4 0 ) Other illegal drugs lifetime use ref 1. 33 (0. 90 1. 96 ) 2.9 4 (2. 23 3.8 8 ) Internalizing behaviors Depression symptoms ref 1 08 (0. 81 1. 44 ) 1.2 2 (0.9 4 1.5 8 ) Anxiety symptoms ref 0 97 (0. 74 1. 29 ) 1. 16 (0.9 0 1. 49 ) Weight concerns ref 1.0 9 (0.60 2 00 ) 2 28 (1. 57 3 30 ) Peer influence Stimulant use by friend ref 2 35 ( 1 78 3. 11 ) 5 63 (4. 12 7 68 ) Note: ADHD Attention Deficit Hyperactivity Disorder; CD Conduct Disorder
57 Figure 3 1. Conceptual diagram of the association between externalizing behavior s and the non medical use of prescription stimulants among youth. Figure 3 2 Conceptual diagram of the moderation model with ADHD as an example Moderator Gender (M) Outcome Non medical use of stimulants (Y) Predictor ADHD diagnosis (X) Moderator Gender Outcome Non medical use of stimulants Predictor Externalizing behaviors ADHD diagnosis CD symptoms
58 CHAPTER 4 AGE OF INITIATI ON AND RISK FACTORS OF NON MEDICAL USE OF PRESCRIPTION STIMULANTS Background Substance use by youth continues to be a national public health problem; the rate of use of illicit drugs including the non medical use of prescription medication s (such as stimulants, sedatives, benzodiazepines ) do es not seem to show any signs of decline (Johnston et al., 2014) About 5 10% of youth in the general population report t he non medical use of prescription stimulants (Conn and Marks, 2014; Cottler et al., 20 13; King et al., 2013; Nargiso et al., 2015 ; Wang et al., 2015 ) However, the non medical use of prescription stimulants and other medications have rarely been included when examining the development of substance use initiation and progression. Substa nce Use Initiation a nd Progression Theories Many studies on substance use among youth indicate the progression of use from licit substances to illicit in a sequential manner known as the Gateway Hypothesis and that the progression to a drug stage is unlikely without first having used a drug that was lower in the drug stage (Kandel, 2002) Although the universality of the theory has been questioned by some resear chers (Golub and Johnson, 2004; Patton et al., 2005) several studies that focused on youth in the United States ha ve shown the progression of substanc e use that begins with alcohol or tobacco not legal among teens and advances to the regular use of the substance and use of other substances such as marijuana, cocaine and methamphetamine (Bracken et al., 2013; Choo et al., 20 08; Chung and Martin, 2001; Degenhardt et al., 2009; Kandel and Yamaguchi, 1993; Kirby and Barry, 2012; Sartor et al., 2013; Tarter et al., 2012) Some studies suggest that the type of substance misuse d can differ by the existing opportunities to access the
59 substance such that drug use laws or socio cultural practices and norms may increase or decrease the availability of certain substances (Wagner and Anthony, 2002) When considering access and obtainability, the last two decades in the US has witnessed increased recognition of Attention Deficit Hyperactivi ty Disorder (ADHD) and its treatment with pharmaceutical stimulants which has played a major role in the environmental availability of these medications (Gillespie et al., 2013) Also, the majority of the youth indicate that their non medical use of prescription stimulant s is largely supplied by friends or family with prescriptions (Conn and Marks, 2014; Cottler et al., 2013; Lasopa et al., 2015; Nargiso et al., 2015) Another body of research has demonstrated that not all youth progress through the substance use sequence from licit to illicit d rugs Such deviations occur specifically w ith marijuana use being reported more than alcohol use (Johnston et al., 2012) (1977) posit the Problem Behavior Theory which suggests that substance use and abuse are symptoms of a common vulnerability to co varying problem behavio rs (such as substance use, behavior problems, delinquency and high risk sexual activity among others) which are associated with a common set of risk factors. In accordance with the Problem Behavior Theory, prior research has shown that early use of licit s ubstances (tobacco and alcohol) increases the risk of illegal substance use including the non medical use of stimulant medications as well as related psychosocial problems such as violent behaviors and drunkenness, (Gruber e t al., 1996; McCabe and West, 2013; Moss et al., 2014; Olthuis et al., 2013) self harm (Zha, 2009) traffic accidents (Heeren et al., 2002) and loss of social relationships (Gruber et al., 1996)
60 Whether substance use occurs in a sequential progression or with other problem behaviors, the risk of adverse physical and mental health consequences is seen with e arly substance use (Behrendt et al., 2009; Brook et al., 2007; Gil et al., 2004; Henry et al., 2012; King and Chassin, 2007; Peleg Oren et al., 2009 ) The early onset of non medical use of prescription stimulants has been linked to the development of later dependence (McCabe and West, 2013) Even though for some youth, non medical use of prescription stimulants may last for on ly a brief period of experimentation, use of these substances in youth do have serious adverse consequences as demonstrated by a dramatic rise in emergency department visits related to the misuse of stimulants (Substance Abuse and Mental Health Services Administration, 2014) Externalizing Problem Be haviors a nd Risk of Initiating the N on medical Use of Stimulants Several factors associated with the initiation of substance s such as alcohol, tobacco and other drugs have been studied that includes as family history of substance use, genetics, inconsistent parental practices, different opportunities for access to substances and peer influence E xternalizing problem behaviors such as inattention, hyperactivity and aggression have been found to be particularly important in substance use initiation (Grant and Dawson, 1998; Levy, 2014; Molina et al., 2013, 2007; Mrug et al., 2010; Nelson et al., 2015; Pa gan et al., 2006; Palmer et al., 2010; Prescott and Kendler, 1999; Wagner and Anthony, 2002) Consistent with other substance use literature, a small number of studies suggest a positive relationship between externalizing behavior problems such as CD an d ADHD symptoms and the onset of NMU of stimulants (Arria et al., 2011; Cassidy et al., 2013; Dussault and Weyandt, 2013) While using stimulants for ADHD treatment itself does not increase the risk of
61 NMU, many studies have found an increased risk of substance use among those diagnosed with ADHD or having symptoms of ADHD ( Pittsburg et al 2013 ). Having both ADHD and CD seems to confer greater risk of substance use at younger ages among adolescents (Brinkman et al., 2015) Two studies investigated the role of conduct problems in the association between early ADHD and later NMU of stimulants (Brook et al., 2014; Van Eck et al., 2013) Both studies found that ADHD and CD were independently and directly associated with the NMU of stimulants. Further Brook and colleagues (2014) in their 22 year longitudinal study of 14 year old youth s found that in addition to the direct path between ADHD and NMU of stimulants ADHD was also indirectly linked to NMU of stimulants through CD problems and other substance use disorder s In another study, CD symptoms increased th e risk of initiation across all substance types at the age of 15 years; moreover, continued risk of initiating till 21 years of age was observed for stimulants but not for all other substance types (Hopfer et al., 2013) In other studies on substance use among youth not specific to stimulants there are some indications of the role of gender in e arly substance use involvement specifically the use of tobacco increase d illegal drug use among boys (Baumeister and Tossmann, 2005; Guxens et al., 2006) A nother study showed that girls transitioned from experimental tobacco use to regular use faster than boys (Galanti et al., 2001) While there has been increased research interest on the epidemiology of NMU of stimulants among youth, less attention has been given to the initiation of the non medical use or on factors that influence the initiation and consequences of early non med ical use Given the potential adverse effects associated with the NMU of stimulants knowledge regarding the age of initiation into non medical use of prescription stimulants
62 will be particularly important for efforts that intervene with the purpose of pre vention or delay of initiation and long term use. Further, the ascertainment of modifiable risk factors such as externalizing problem behaviors of ADHD and CD symptoms that may influence onset of non medical use is important with a view to specialized plan ning of prevention strategies and design of early detection programs based on the risk factors Current Study There have been a number of cross sectional studies that report on the epidemiology of NMU of stimulants among youth that have identif ied a number of associated potential factors; however a dearth of prospective studies that examine the trajectory of NMU of stimulants in youth is observed. Ideally, prospective studies in the general population can be used to examine the initiation into NMU of stimulants and factors that increase the risk of onset. However, these approaches are costly, prone to attrition and take long to follow up. An alternate approach is to use cross sectional survey data and construct pseudo longitudinal observations draw ing on the retrospective age of onset of NMU of stimulants assessed. Survival analysis can then be used to model time to NMU of stimulants allowing the examination of factors associated with the transition from a non use state to one of NMU of stimulants T his method has been used previously to assess the age of onset of other substances such as cigarettes, alcohol and marijuana (Butter worth et al., 2014; Kalaydjian et al., 2009) Here again, majority of empirical studies o n the initiation and development of substance use including NMU of stimulants have focused on the sequencing of drug use and ages of onset as the parameters of impo rtance with comparatively less focus on factors that influence onset
63 In the present study, survival analysis of a data from a nationally representative sample of youth aged 10 to 18 years in the US was used to examine: 1) the period ( in terms of age ) when youth are at highest risk for initiation of non medical use of prescription stimulants; 2) the role of externalizing problem behaviors of ADHD and CD symptoms in the initiation of non medical use of prescription stimulants; and 3) whether gender differenc es in the initiation of non medical use of prescription stimulants in youth exist. Additionally, we ascertained the impact of prior use of other substance s (such as cigarette s alcohol and marijuana) on the initiation of NMU of stimulants The aims of the study were addressed with the following hypotheses: 1) self reported ADHD diagnosis in youth will increas e the hazards of initiation of NMU of stimulants ; 2) h azards of NMU of stimulants initiation will be higher among youth with self rep orted CD symptoms; 3) there will be gender differences in the initiation of NMU of stimulants with males showing increased hazards o f initiation ; and 4) among non medical users of s timulants, the use of alcohol, cigarette, and marijuana prior to non medica l stimulant use will be associated with increased risk of N MU initiation compared to later users of other substances Methods The paper reports on analyses conducted on data from the National Monitoring of Prescription Stimulants Study (N MAPSS). The N MAP SS is a cross sectional study of 11,048 youth between the ages of 10 and 18 years who were interviewed about their knowledge, use and misuse of prescription stimulant s from ten cities across the United States. Data collection was carried out in four cross sections from fall 2008 to spring 2011. Other details of the study are reported elsewhere (Cottler et al., 2013) The study
64 was approved by the Institutional Review Boards of Washington University in St Lo uis for the first three cross sections and the University of Florida f or the fourth cross section Sample The present analysis is restricted to youth in the second, third and fourth cross sections (n=8330); the first cross section of the N MAPSS did not assess the age of first use of stimulants among youth and was therefore excluded from this study. Further, we excluded youth on certain criteria specific to this study described below. In clusion criteria Youth who were recruited in the third and fourth cross section of the study with lifetime non medical use of stimulants and reported their first age of stimulant use were included in the present analyses. Exclusion c riteria Youth with reported stimulant use but with missing age of onset were excluded from the analysi s. Those youth who met the criteria for medical use only (MU) of prescription stimulants defined as use of stimulants that were prescribed by a psychiatrist or doctor were not included in the analyses. Additionally, youth who met criteria for NMU but who a as excluded. Based on the exclusion criteria, we did not include 434 youth who met the definition of MU only because we aimed to assess risk of first non medical use into prescription stimulants relative t o non users. A dditionally 201 youth who met criteria for MU only but also responded positively to any of the questions that indicated NMU were not included in the present study. This was because youth were only asked their age of first stimulant use and w ere not specifically asked when they started either NMU or MU
65 only ; distinguishing whether NMU was prior or later to MU only was not possible Further, four youth s who did not report their age of first stimulant use and one reportin g first use of NMU at two years of age were excluded from further analyses. Following removal of youth based on the exclusion criteria, the final sample for this study consisted of 7691 youth. Measures Event measure The change from non user to NMU of stimulants was the main outcome or event of stimulants was defined as the u se of any of the 5 stimulants: Adderall or Adderall XR Concerta Ritalin Daytrana and Vyvanse which belonged to someone else (parents, brother or sister, different family member, someone from school or from work, someone unknown and other) or ingested via non oral routes (such as snorted or sniffed, smoked), or were used for Youth who met the definition of NMU of stimulants w ere censored at the age of interview. Age of non medical s timulant u se initiation The N MAPSS assessed youth for lifetime and past 30 day use of prescription stimulants; youth who had ever used any of the five stimulants were asked to report on their age at first use for each individual drug. The dependent variable of interest in this analysis was the age of NMU of stimulants initiation defined as the self reported age in years at which youth had first used any of the five stimulant medications assessed for in the N MAPSS.
66 Length of f ollow up Because cross sectional data was used, we construct ed pseudo longitudinal observations drawing on the retrospective age of onset assessed to m odel time to NMU of stimulants The difference in reported first age of onset of prescription stimulant use from the current age was used to ascertain time taken to develop NMU of stimulants across the length of follow up time. The length of follow up was the age at survey fo r those who had not initiated NMU of stimulants This allowed the tracking of age of NMU of stimulants initiation over time and was assessed as an increase of 1 point for 18 years (maximum year possible) thereby creating pseudo longitudinal data necessary for survival analysis. ADHD ADHD w as measured Has a doctor ever told you or your parents that you have Attention Deficit Disorder (ADD) or ADHD ? Youth who responded considered to have ADHD while those who were considered no ADHD CD s ymptoms warnings have you r variables were coded 1 when youth responded having at least one episode having received tickets or warnings, or being arrested at least once, and coded 0 when there
67 were none. Based on these items, a variable to represent conduct problems was created when youth positively responded to a minimum of three of the items. Other su bstance use i nitiation To assess th e effects for prior use of other substance s on NMU of stimulants measures of cigarette, alcohol and marijuana initiation were included. Age of initiation of cigarette, alcohol and marijuana use was assessed among youth who positively responded to having u sed the three substances in individual questions. We defined age (in years) of initiation of cigarette use as the age a participant reported they first engaged in cigarette use Age of initiation of alcohol use was defined as the age (in years) a responden t reported they first had a full alcoholic drink. Age (in years) of initiation of marijuana use as the age a respondent reported they first used marijuana. Because we were interested in assessing the impact of substance use on NMU of stimulants, we created a three level variable s for cigarettes, alcohol and marijuana use to represent: 1) use of each of the substances prior to NMU of stimulants among those with NMU of stimulants ; 2) use of each of the substances later than NMU of stimulants among youth with NMU of stimulants and 3) a third level to indicate no NMU of stimulants (which also included youth who used any of the three substance s but did not engage in NMU of stimulants) Covariates Gender and Caucasian race coded as dichotomous variables were inc luded as covariates in all models. Area of residence was included as a three level variable with urban suburban and rural designations based on zipcode level information from youth.
68 Statistical A nalyses Descriptive analyses were conducted to present cumulative incidence proportions of non medical use of prescription stimulant medication. Given that an individual X years of age cannot report the NMU of stimulants initiation greater than X the survival mod el offers a fairly robust alternative to analyze such data (Chen and Unger, 1999) .The analyses tested the hypothesis that: 1) hazards of NMU of stimulants initiation would differ by externalizing behavior probl ems mainly ADHD status and CD status; and 2) prior use of alcohol tobacco or marijuana would increase hazards of NMU of stimulants initiation compared to later use Discrete time survival analysis was used to model the age in years at first use which mode ls the conditional probability that a particular youth will experience an event (NMU of stimulants ) in a given period of time (age). The hazards in this analysis represent the risk that a youth who has not yet initiated NMU of stimulants at the beginning o f a given age will initiate NMU of stimulants at any given time within the given age period. This is a robust technique to investigate the time taken to develop NMU of stimulants across the length of follow up time. The Hazard X time or H(t) therefore esti mates the probability of initiating NMU of stimulants by age The SAS Proc Lifetest was used to estimate the hazards of NMU of stimulants initiation. Hypothesis 2 tested whether the estimated hazards H(t) of NMU of stimulants initiation differed by ADHD and CD status and prior to NMU of stimulants onset of cigarette, alcohol and marijuana using the Strata option and log rank tests. Thereafter, the Cox proportional hazards survival models assessed the influence of ADHD, CD, prio r to NMU of stimulants use of substance (cigarette, alcohol, marijuana) and other covariates (gender, race residence ) on the hazards of NMU of stimulants onset. The Cox model
69 does not make assumptions about the form of the baseline model akin to being non parametric. However, two important issues were considered prior to the application of the model results. First, the N MAPSS design ensures that the mechanisms used to censor youth were not associated with the probability of occurrence of NMU of stimulants (event); the follow up period is in fact dependent on the age of interview. Therefore, the assumption of n on informative censoring is satisfied. An important second assumption in the Cox model is that of proportional hazards wherein survival curves for tw o strata (in this analyses values for ADHD, CD and gender variables) are required to have a constant relative hazard function (i.e. the hazard functions are proportional over time). This constant relative hazard comparison by ADHD and CD status and gender were evaluated graphically using the "log log" plots which indicated that the proportional hazards assumption was met. The SAS Proc Phreg assessed the hazards of first non medical use of prescription stimulant medication initiation as a function of ADHD an d CD status while adjusting for covariates of gender, race and area of residence. The model also tested hypothesis 2 whether the use of other substances (cigarette, alcohol, marijuana) prior to NMU of stimulants would increase hazards of NMU initiation. B ivariate and adjusted hazard ratios (HR) with 95% confidence intervals (CI) are reported. Results C haracteristics o f Youth with N on m edical U se o f Prescription S timulants Table 4 1 displays the characteristics of youth in the current analyses. Of the entire sample (n=769 1 ), 6.07% of youth reported the NMU of stimulants The prevalence of the self reported externalizing problem behaviors of ADHD and CD symptoms w ere similar at 10.6 % and 10.3% respectively. However, the rate of ADHD
70 was more than 11 time s higher among those with NMU of stimulants compared to non users of stimulants The rate of CD symptoms was about two times higher among youth with NMU of stimulants relative to youth who did not use stimulants. Youth with NMU of stimulants were more like ly to be male, Caucasian and less likely to live in rural areas. As shown in Table 4 2, almost half of youth with NMU of stimulants had ever used alcohol (48.2%) with about 15% report ing the initiation of alcohol use before the NMU of stimulants. Approximately one third of youth with NMU of stimulants ha d ever used cigarettes or marijuana with almost 12% reporting initiating either of the two substances prior to NMU of stimulants. Overall, higher proportions of you th reported u se of alcohol, cigarette s and marijuana after the non medical use of stimulants. Hazard o f Initiating N on m edical Use of S timulant s The hazard estimates for non medical use of prescription stimulants are presented in Table 4 3 as a life table. Overall, the estimated hazard of NMU of stimulant s was low ( 2% or less) prior to age 10 and increase d to about 3.5% at age 11 years. From 1 2 years of age, the hazard of NMU of stimulant s initiation steadily increase d by about 0.6 0.8% with each year increase in age till age 18 years when the hazard estimate reache d the highest at 7.6% (also presented graphically in Figure 4 3 ) The hazard estimates of initiation into non medical use of stimulants differe d by the externalizing behaviors of ADHD, CD symptoms and also by gender. As shown in Figure 4 2 at every age, the h azard s of NMU of stimulants initiation is markedly higher among youth with ADHD compared to those without ADHD (Log rank 2 (1) = 241.2838, p < .0001). Similarly, F igure 4 3 indicates that y outh with CD ha d significantly increased hazard s of initiating NMU of stimulants relative to non CD youth (Log rank 2 (1) =
71 31.4410, p < .0001). A gender difference in the hazards of NMU of stimulants initiation was also observed males had higher hazards of initiating NMU of stimulants at every age assessed compared to females ( Figure 4 4 ; Log rank 2 (1) = 7.9608, p < .0048). Effect of Externalizing Problem Behaviors o n t he Initiation o f Non m edical Use o f Prescription Stimulants As shown in Table 4 4 in multivariable analysis, the externalizing behaviors of ADHD and prior use of cigarettes, alcohol and marijuana increased the hazard risk of NMU of stimulant initiation. I n the initial unadjusted model, youth with history of ADHD s howed about 4 times the risk of initiating NMU of stimulants compared to those without a history of ADHD ((hazards ratio (HR): 4.0 2 95% confidence interval (CI): 3.32 4.86)). When controlling for gender, race, area of residence, CD symptoms, and prior use of alcohol, cigarettes and marijuana use, the hazard of initiating NMU of stimulants among youth with ADHD decreased only slightly with HR at 3.82 However, report ing CD symptoms did not increase the hazard of initiating NMU of stimulants when adjusting f or other variables in the model. The hazards of NMU of stimulants were markedly high in the u nadjusted model s comparing risk from prior alcohol, cigarette and marijuana use to later use. When adjusting for other variables, the risk of NMU of stimulants wa s higher among youth with prior use of cigarettes ( HR 3.13 ), marijuana ( HR 2.22) and alcohol ( HR 1.39 ) compared to youth with non use of the substances although the hazard estimates were lower compared to the unadjusted models Effect o f C ovariates o n the Initiation of Non m edical Use of Prescription Stimulants When adjusted for all other variables, t he hazard s of initiating non medical use of prescription stimulants did not differ by gender, race or residence Interaction terms
72 between ADHD and gender on the risk of NMU of stimulants onset were tested, however, it was not significant suggesting that ADHD only independently influence s the initiation into NMU of stimulants Sensitivity A nalysis To ascertain the impact of not including medical users only and both medical users with non medical use in the final model presented in the results, sensitivity analysis w ere conducted. First, a model with MU only considered as non users was compared to the final model t hat excluded the group of youth. Second, results for a model with both MU and NMU considered as non users was compared to the final model that excluded the group of youth. Third, results for the model with both NMU and MU considered as NMU and MU only considered as non users were compared to the final model. Results indicate no significant differences in the parameter estimates (HR) for the exposures of interest i.e. ADHD and gender on the hazards of NMU of stimulants initiation in the fina l model when compared to the other models used in the sensitivity analysis. The results of the sensitivity analysis are presented as a table in Appendix B Discussion Using survival analysis methods, we sought to examine the ages of highest risk for first non medical use of prescription stimulants and factors that influence the initiation among youth aged 10 to 18 years in the US. In this study, w e found that the hazard of non medical use of stimulant s initiation amplified with each year increase in age. A lthough the hazard of non medical use of stimulants was fairly low before 10 years of age, it increased each successive year after age 11 and peaked at 1 8 years of age when our study period ended This finding suggests that the period of 11 to 18 years of age is a critical period for NMU of stimulant prevention among youth in the US.
73 As hypothesized, the hazard of initiating non medical use of stimulants differed by whether youth reported externalizing behaviors of ADHD symptoms but this was not true for CD symptoms. Prior use of alcohol and marijuana use also increased the hazard s of initiating non medical use of stimulants ; this association was not observed for prior cigaret te use. Our findings are consistent with the few studies that examine the age of onset of non medical use of prescription stimulant medication. For example, Austic (2015) used data from a national household survey and reported on incidence rates of non medical stimulant use for the years 2004 to 2012. The study found the peak ages for non medical stimulant use were between the 16 and 19 years of age and the risk for initiating stimulant non medical use was marginally lower in the age ranges 12 to 15 years and 20 to 21 years. Further, youth differed in the risk of non medical use of stimulants initiation by race; Whi tes or Native Americans showed higher incident non medical use of prescription stimulants at each individual year of the study. In another study by Austic and others (2015 ) the highest incidence of initiation of non medical use of prescription medication (analgesics, stimulants, sedatives and anxiolytics) was at age 16 among youth in the Detroit area. Although the methodology used in our study differs from th ose of the Austic stud ies overall results seem to indicate that preventive efforts for non medical use of prescription stimulants should begin by the early teen years. In related work, McCabe and others (2007) found app roximately one fourth of the 13 % prescription stimulant dependent youth began use before 13 years as compared to less than 10% of those who began use after age 21. Moreover, the association of early non medical use with increased risk for later dependence was more likely to occur among
74 prescription stimulant non medical users as compared to any other type of prescription drug class. These findings suggest that different patterns and risk factors may exist depending on the type of drug abused which further justifies our attention to prescription stimulant non medical use and its onset among youth. Our findings suggest that a factor tha t influence s the initiation into non medical use of prescription stimulants is the externalizing behavior of ADHD which seem s to independently increase the risk (Hopfer et al., 2013) There are some possible explanations for an association between ADHD and non medical use of prescription stimulants It has been suggested that early initiators of substance use self select and may already be different from their peers at the time of subst ance initiation A mong non medical users of prescription stimulants, ADHD could be involved as a self selection factor. P revious work has shown that compared to non users, higher likelihood of attention problems were reported by students who used stimulant s non medically (Rabiner et al., 2009) Although conducted on college students, related to our findings is the study by Arria and others (2012) which found that self reported ADHD symptoms were predictive of continued non medical stimulant use over a period of four years. These and prior findings suggest that some youth may engage in the non medical use of stimulants as a way to self treat their perceived or untreated ADHD symptoms (Rabiner et al., 2013) On the other hand, non medical use of stimulants may be used to achieve academic success even when youth do not have symptoms of ADHD. Available evidence does indicate that youths reasons for substance use differs based on the class of medications and their perceived effects and stimulants are known to impr ove attention and concentration among those with ADHD ( Boyd et al., 2006; Bogle and
75 Smith, 2009 ) Further, ADHD has been linked with increased impulsivity which is associated with impairments in executive functioning and these in turn are associated with p oor decision making that result in the use of substances (Dom et al., 2006) Although we expected that CD symptoms would be important in the initiation of non medical stimulant use our findings did not support this hypothesis. Other studies have found an association between CD and elevated risk of initiation for non medical s timulant use (Brook et al., 2014; Hopfer et al., 2013) One possible reason for the non significant findings in the present study could be related to the stronger effects of prior use of other substances (such as ciga rettes, marijuana and alcohol) on the initiation of NMU of stimulants compared to CD Overall, our study findings are in accordance with the Problem Behavior Theory and a large body of research that suggest s that non medical use of stimulants is part of a generalized problem behavior that exposes youth to a variety of co varying risk behaviors of other substance use (such as alcohol and marijuana) and externalizing behaviors of ADHD (Jessor and Jessor, 1987; Lerner et al., 2013; Molina and Pelh am, 2003; Morral et al., 2002) The risk of non medical use of prescription stimulants has been shown to be highly associated with other substance use (cigarette, binge drinking, marijuana, other illicit drugs and non medical use of other prescription d rugs) among a national sample of youth (McCabe and West, 2013) To some extent, our findings are also in line with the Gateway Hypothesis (Kandel, 2002) we found the prior use of cigarettes, alcohol and marijuana served akin to gateway subst ance s and increased the hazards of initiation into non medical use of prescription stimulant medication.
76 Collectively, there is indication that the onset of non medical use of prescription stimulants among youth seems to occur quite early in the developmental period and that prior use of alcohol and marijuana use app ears to increase the risk, it may be worthwhile to include prevention efforts for non medical use of stimulants when those for other substance use are carried out. T he delay of initiation of overall substance use may have beneficial effects by reducing the time in a formative period of the life span when an individual experiences the deleterious effects of drugs. While our findings reinforce the importance of preventi ve efforts to delay the age of onset of prescription stimulants or prevent use entirely as part of substance use prevention programs in general, it is noted that several drug education and prevention programs in school both at the middle and high school le vels have not been very successful when the salient risk factors have not been taken into account (Griffin and Botvin, 2010) Specific screening programs for poorly controlled ADHD symptoms may be incorporated as measures to identify those at risk for non medical use of stimulants. Though the present study offers important insights into the understudied area of initiation in to non medical use of prescription stimulants among youth, it is not without limitations. The cross sectional nature of the design did not allow the inclusion of some important variables that could have allowed us to infer causality based on actual tempora lity. However, care was taken to include in the analyses only variables for which the independent variables preceded the onset of non medical use with a reasonable degree of certainty, strengthening the inferences made. We focused our analysis on non medic al use of stimulant s even if youth only used stimulants once and diagnostic criteria for abuse or dependence o n stimulants were not used T here may be important
77 differences in youth who non medically use stimulants once and those who go on to develop disorders. Further, we used retrospective reports of age at first use which may substance use onset as being more earlier or remote th an that of another substance (Johnson and Schultz, 2005) Lifetime use measures of substance use may be biased by recall errors due to fading of memory over time and particularly when age of onset is farther apart in time from the curren t age. Despite the limitations noted, this study provides important insight s into the nature of initiation into non medical use of prescription stimulants in a representative national sample of youth. The study was also able to identify some of the importa n t factors involved in increasing risk of first non medical use such as ADHD symptom s Given that longitudinal studies are not always feasible, we were able to model the hazards of initiation into non medical use of prescription stimulants using cross sect ional data. Findings of the study with respect to the period of highest risk for onset has important implications for the design of programs that aim to either prevent or delay the age of first non medical use of prescription stimulant s
78 T able 4 1. Characteristics of youth aged 10 to 18 years by lifetime prescription stimulants use status. Characteristic No use (n=7228) Non medical use (n=467) p value Total (n=7691) Male gender 45.6 % 53.1 % 0.0017* 46.1 % Caucasian 40.1 % 60.6 % <.0001*** 41.3 % Residence Urban 48. 6% 38.1 % 47.9 % Suburban 36.8 % 40.9 % <.0001*** 37.0 % Rural 14. 7% 21.0 % 15.1 % ADHD 6 6 % 7 3. 7 % <.0001*** 10.6 % CD symptoms 9.7 % 18.9 % <.0001*** 10.3 % p <.05, ** p <.0 0 1, *** p <.00 0 1 Note: ADHD Attention Deficit Hyperactivity Disorder; CD Conduct Disorder Table 4 2 Life time s ubstance use among youth with non medical prescription stimulants use (n=467) Substance type Non medical use of stimulants Alcohol use No use 5 1.0 % Onset before NMU 15.2 % 33.0 % Cigarette use No use 64.5 % Onset before NMU 11. 6% 22.7 % Marijuana use No use 61.0 % Onset before NMU 11. 8% 26.3 % Table 4 3 Life table estimates of hazard probabilities of initiation into non medical use of stimulants among youth (N= 7691 ) aged 5 to 18 years from 2009 to 2011. Age Number entering interval Number censored Number of terminal events Hazard Standard error Cumulative Hazard Standard error 5 years 7691 0 0 0 0 0 6 years 7691 52 52 0.00676 0.000934 0.00676 0.000938 7 years 7639 88 36 0.0114 0.00121 0.0115 0.00122 8 years 7603 125 37 0.0163 0.00144 0.0163 0.00146 9 years 7566 158 33 0.0205 0.00162 0.0207 0.00165 10 years 7533 225 67 0.0293 0.00192 0.0296 0 .0 0197 11 years 7259 266 41 0.0347 0.00209 0.0352 0.00216 12 years 6920 324 58 0.0428 0.00233 0.0436 0.00243 13 years 6427 397 43 0.0492 0.00251 0.0503 0.00263 14 years 5584 427 30 0.0543 0.00266 0.0266 0.0557 15 years 4586 448 30 0.0605 0.00287 0.0622 0.00305 16 years 3387 456 21 0.0663 0.00313 0.0684 0.00334 17 years 2156 463 8 0.0698 0.00335 0.0721 0.00359 18 years 1044 463 7 0.0760 0.00407 0.0788 0.00439
79 Table 4 4 Estimated hazards of non medical prescription stimulant use initiation among youth 10 to 18 years of age in the N MAPPS Characteristic Unadjusted HR (95% CI) Adjusted HR (95% CI) Gender Female r ef r ef Male 1.2 8 (1.07 1.54) 1. 11 ( 0.92 1. 34 ) Race Non Caucasian r ef r ef Caucasian 2.2 4 (1.8 5 2.7 0 ) 1.18 ( 0 97 1.43 ) Residence Suburban vs. urban 1.42 (1.1 6 1.7 5 ) 1.2 1 ( 0 98 1. 49 ) Rural vs. urban 1.7 7 (1.3 8 2.2 8 ) 1. 08 ( 0 89 1. 49 ) ADHD diagnosis No r ef r ef Yes 4.0 2 (3.32 4.86) 3.82 ( 3.00 4.85 ) CD symptoms No r ef r ef Yes 1.9 4 (1.5 3 2.4 4 ) 0.93 ( 0 73 1. 18 ) Cigarette use Onset after NMU r ef r ef 13.98 ( 12.50 15.29 ) 3.13 ( .2.21 4.43 ) Alcohol use Onset after NMU r ef r ef 12.94 ( 11.72 14.29 ) 1.39 (1. 06 3 09 ) Marijuana use Onset after NMU r ef r ef 13.5 1 ( 12.23 14.92 ) 2.22 ( 1.59 4.85 ) Note: ADHD Attention Deficit Hyperactivity Disorder; CD Conduct Disorder ; NMU Non medical Use of Prescription Stimulant s
80 Figure 4 1. Hazard estimates of age of initiation into non medical use of prescription stimulants among youth 10 to 18 years of age. Figure 4 2. Hazard estimates of age of initiation into non medical use of prescription stimulants by A ttention D eficit H yperactivity D isorder (ADHD) status among youth 10 to 18 years of age.
81 Figure 4 3. Hazard estimates of age of initiation into non medical use of prescription stimulants by C onduct D isorder (CD) status among youth 10 to 18 years of age. Figure 4 4. Hazard estimates of age of initiation into non medical use of prescription stimulants by gender among youth 10 to 18 years of age.
82 CHAPTER 5 NON MEDICAL USE OF PRESCRIPTION STIMULANTS AND THE USE OF OTHER SUBSTANCES AMONG YOUTH : A LATENT CLASS ANALYSIS Background A number of studies have indicated that prescription stimulants are now fairly common substances of misuse among youth in the United States; past one month non medical use of prescription stimulant medication (NMU) rates between 5 10 % have been reported in both regional and national studies (Cottler et al., 2013; McCabe and West, 2013 ; Wang et al., 2015 ) Adverse events leading to emergency department visits linked to NMU of stimulants are also at a significant high compared to most other drugs of abuse (Substance Abuse and Mental Hea lth Services Administration, 2013b) As NMU of stimulants continues to be a problem among youth, it is as important to view that NMU of stimulants may reflect more than an isolated form of substance use behavior (McCabe et al., 2014) Youth continue to misuse a number of substances such as alcohol, tobacco and other illicit drugs or use a combination of substances, and it is generally uncommon for youth to use only one particular class of drug (Gilreath et al., 2014) Polysubstance use, increased rates of o ther substance use and high risk behaviors are well documented among youth with NMU of stimulants (McCabe et al., 2012; Sweeney et al., 2013) There has been an emphasis in both the timeliness and need to better characterize contemporary patterns of substance use and abuse among youth with studies that include prescription medications (such as stimulants, opioids, benzodiazepines) together and separately from other drugs of abuse. Much of our understanding of NMU of stimulants and other substance use behavior in youth along with their associated risk factors has primarily been achieved using a variable centered approach. Examples of this approach are many and
83 distinguish specific patterns of substance use among individual substances such as alcohol, tobacco or use across a range of substances; individual substance use variables are examined assuming sample homogeneity (McCabe et al., 2012) In contrast, person centered approaches such as Latent Class Analysis (LCA) assume sample heterogeneity and are used to empirically identify distinct subtypes or classes of individuals based on their shared patterns of observed indicator variables such as types of subs tance used or motives for use (Cleveland et al., 2010a; Lanza and Rhoades, 2013; McCabe and Cranford, 2012) Using LCA, a few studies have suggested considerable variability within the overall subgroup of you ng sub stance users which includes NMU of stimul ants and other prescription medication such as opioids (Cranford et al., 2013) For instance, there are at least some subgroups that consist of polysubstance users, a subtype that engages in alcohol use with non medical use of medication that includes stimulants grouped together (Chen et al., 2014; McCabe et al., 2009) A few studies have characterized youth based on their motives and routes for non medical use indicating that there are motivational subtypes of non medical use of prescription medication and have identified a subgroup of youth who use sti mulants to achieve or perform better (McCabe et al., 2009) Further, important differences in risk factors among subtypes of substance using youth have been shown to vary depending on subtype characteristics such as the number and category of drug s misused (Hall et al., 2010) Despite reports of coexisting use of NMU of stimulants and other s ubstances of abuse, it has been suggested that some risk factors may be unique to NMU of stimulants (Schepis and Krishnan Sarin, 2008) However, a number of prior studies have implicated that shared risk factors for
84 alcohol, tobacco and other drugs exist which include individual, personality an d family related variables (Buu et al., 2009; Harden et al., 2008; Kendler et al., 2003; Thatcher and Clark, 2008; Ystrom et al., 2013) Whether risk factors for NMU of stimulants are indeed different or similar from those associated with multiple substance use is still not entirely clear. Most prior research limited the range of substances studied to alcohol, tobacco, and cannabis with the exception of a few recent studies that include the non medical u se of prescription drugs (Gilreath et al., 2014) O ne recent study conducted in two Midwestern cities sampled youth 12 to 18 years of age from an urban primary care setting and identified a three class model of lifetime substance use (i.e., low substance users, cannabis users, and polysubstance users which included prescription medication misusers) (Bohnert et al., 2014) Because a majo rity of the previous studies that took the LCA approach to examine youth NMU of stimulants and other substance use have utilized student focused samples whether regional or national ( Boyd et al., 2006; McCabe et al., 2009; McCabe and Cranford, 2012) there seems to be a dearth of studies that use nationally representative samples. Given that prescription medication misuse continues, we emphasize the importance of characterizing yout h based on their patterns of substance use that also include the non medical use of prescription medications with representative samples. The generation of generalizable data will be critical in informing tailored intervention efforts to address specific s ubstance use patterns among youth. We had the opportunity to address these gaps in current substance abuse literature using data from the National Monitoring of Prescription Stimulants Study (N MAPSS) which sampled youth as young as 10 years of age from t he general population
85 and assessed a range of substances that include alcohol, tobacco, marijuana, other illicit drugs, non medical use of prescription stimulants, opioids and benzodiazepines The present study aimed to use a person centered approach to id entify empirically determined subtypes of youth based on their past 30 day substance use patterns and examine d socio demographic, behavioral and mental health characteristics of these subtypes using a variable centered approach We test ed whether the class with the most number of substance use types w ould have the highest likelihood of mental and other behavioral problems. Thus, we use d variables across several domains of risk by building on prior research on risk factors for substance use. Methods Sample a nd P rocedures The present study sample was drawn from the larger N MAPSS study conducted across ten cities in the United States using a cross sectional design. Trained interviewers recruited youth 10 to 18 years of age at selected youth friendly entertainm ent venues (such as shopping malls, parks, sports and recreation centers, movie theatres) in four cross sections from 2008 to 2011. Interviews were conducted in private areas of the venues to ensure confidentiality. Youth gave verbal assent for participati on in the s tudy and were provided with a $ 10 at completion of the interview. Study protocols were approved by the Institutional Review Boards at Washington University in St Louis for the first three cross sections and University of Florida for the fourth cross section.
86 Measures Substance u se The N MAPSS survey assessed past 30 day use of seven ty pes of substances which are: 1) Use of the prescription stimulants Adderall or Adderall XR Concerta Ritalin Daytrana and Vyvanse measured individually. Pictures of each type of stimulant medica tion were presented and youth were asked to identify the stimulant by brand name and formulation. Specific questions asked for: a) past 30 day use of each stimulant; b) use of stimulant in greater quantity than prescribed; c) use that belonged to someone e lse (parents, brother or sister, different family member, someone from school or from work, someone unknown and other); d) route of ingestion of stimulant (by mouth, snorted/sniffed, smoked and other), and e) reasons for prescription stimulant use. Questio ns also assessed if each stimulant endorsed for use had been prescribed by a psychiatrist or a doctor. For the present study, past 30 day NMU of stimulants is defined as the use of any of the five aforementioned stimulants for use other than by mouth (exce pt Daytrana which comes as a patch ), use of any stimulant that belonged to someone else or use that was more than prescribed. Youth who met this definition were considered as non medical p rescription stimulant users. 2) Use of the prescription opioids Vic odin and Oxycontin were assessed. Pictures of the medications were presented and youth were asked whether they had used any in the last 30 days. Youth who responded positively to the use of any of the two prescription opioids by non oral routes (snorted/ sniffed, smoked or other) or use that belonged to someone else were categorized as non medical prescription opioid users. 3) Use of prescription benzodiazepines Valium and Xanax were measured. Reporting the use of any of the two prescription benzodiazepi nes assessed via non oral routes or use that belonged
87 to someone else indicated classification into non medical prescr iption benzodiazepine users. 4) days, on how many days did the past 30 days was regarded as alcohol users while those without any alcohol use in the past 30 days even if they were lifetime users were categorized as non users of alcohol in the past 30 days. 5) Cigarette use in the past 30 days was measured among asked if they still s moked every day, on some days or no use; those who reported smoking every day or on some days of the last 30 days were considered as cigarette users. Youth with no cigarette use in the lifetime question and lifetime cigarette use but with no use in the pas t 30 days were categorized as non users of ci garette in the past 30 days. 6) marijuana on at least one day in the past 30 days was categorized as marijuana users while no use in the past 30 days or lifetime q uestion were categorized as non marijuana users. 7) Other illicit drug use measured in the N MAPSS survey included cocaine, heroin, club drugs, ha llucinogens (LSD or mushrooms), anabolic steroids, methamphetamine and inhalants (gasoline or paint). Youth who endorsed at least one of the substances were classified as other illicit drug use users. The distribution of use of all seven substances descri bed above to be used in the study was examined. There were 41 cases with missing data for all the types of substance use variables, and these respondents were excluded from the present
88 analyses resulting in a final sample size of 11,007 youth on whom laten t class analysis was conducted. Socio d emographic c haracteristics Socio demographic variables were assessed in the N MAPSS that include d age, gender, race and ethnicity. Zip code level information was used to categorize area of residence into urban, suburban and rural. The survey also included items that elicited whether youth lived in a dual parent household grades i n school (dichotomize vs. or F sleep timings at 12 am or later and self rated general level of health (excellent, good, fair and poor). Youth were asked the number of meals they ate in a week with family; a categorical variable with less than five meals a week with family was coded 0 and five meals or more were coded as 1. These socio demographic variables were examined to elucidate differences in substance use subtypes. Mental and behavioral h ealth Based on items in the N MAPSS, the following four variables representing mental and behavioral hea lth were created: 1) diagnosis of ADHD by a doctor ; 2) conduct problems (ever got into a lot of trouble at home or at school, ran away from home overnight and used or threatened someone with a weapon); 3) we ight issues (very afraid of gaining weight, tried to lose weight by making oneself vomit, taking pills to lose weight, not eating for a day or two and exercising too much); 4) depressive symptoms (past 12 months loss of interest or depressed for at least 2 weeks); 5) anxiety symptoms (lifetime worry or stress for 6 months or more). ADHD, d epressive and anxiety symptoms were dichotomous variables We also categorized conduct problems and weight issues into dichotomous variables by coding 1 when youth endorse d at least
89 of the individual items for conduct problems and weight issues Peer s ubstance u se yo Analyses The prevalence and distribution of the seven types of subst ance use among youth w ere examined. Using Mplus 7.0 (M uthn and Muthen, 2007) LCA was applied to empirically identify subtypes of youth with similar patterns of substance use that explain their responses based on the observed substance use indicator variables (past 30 day use of alcohol, marijuana, cigare ttes, non medical use of prescription opioids, non medical use of prescription benzodiazepines, non medical use of prescription stimulants and other illicit drugs respectively). The substance use variables were considered as indicator s of the latent classes and it wa s expected that observed data among youth in the same latent class w ould cluster while it w ould differ from that of youth in other latent classes. Model building was carried out iteratively till an optimal model fit was achieved based on f it statistics and interpretability of the subtypes. The fit of the LCA models was evaluated with the Lo Mendell Rubin adjusted likelihood ratio test (LMR LRT), Bootstrapped Likelihood Ratio Test (BLRT), Bayesian Information Criterion (BIC) adjusted Bayesia n Information Criterion (ABIC) and entropy value (Lo et al., 2001; Nylund et al., 2007) The LMR LRT and BLRT examine s whether a model with one additional class (k classes) better describes the data compared to a
90 model with k 1 classes; a significant p value (<0.05) denotes improvement in model fit suggesting the selection of the model with k classes over the k 1 class m odel. The BIC and ABIC indicates the parsimony of the model; smaller values are indicative of an improved model and differences in BIC/ABIC of at least 10 relative to the previous model are taken as an indication to favor one model over another (Raftery, 1995) The BIC and ABIC have been shown to consistently identify the correct number of class es in categorical LCA mode ls (Nylund et al., 2007) The entropy value reflects the classification quality of the model with values closer to 1.00 represent ing increased accuracy in assignment to a class. The conceptual interpretability and size of the classes were also considered in the model selection process. Approximately 2.8% of the respondents had missing data on one or more explanatory variables. Mult iple imputation procedure was carried out assuming missing at random (MAR) using SAS 9.4. Multinomial logistic regression analysis with the imputed data was then carried out to examine the correlates of the LCA subtypes identified which included socio dem ographic characteristics, behavioral and mental health indicators, and peer stimulant use. The adjusted odds ratios (AOR) and 95% confidence intervals (95% CI) are reported. Results Substance U se Patterns Among Youth Substance use in the sample is displa yed in Table 5 1. About 37% of all youth had used at least one type of substance in the past 30 days. The use of any individual substance by youth ranged from approximately 2% to around 2 7 %; the most common s ubstance used was alcohol (27 %) while the least common substance misused was
91 benzodiazepines (2%). Overall, lower rates of all types of non medical use of prescription drugs were endorsed by youth relative to alcohol, tobacco, marijuana and other illicit drugs. LCA Defined Subtypes o f S ubstance U se a mong Youth Following best practices in LCA, the identification of the best fitting model was based on statistical fit measures as well as interpretability of the models. As seen in T able 5 2, t he LCA of the seven observed substance use var iables indicated that a 2 class model (BIC 38705.971; ABIC=38658.303; entropy=.85) differed from a 1 class model (VLMR p<0.01; BLRT p<0.01). Next, the 3 class model (BIC 38252.469; ABIC=38179.378; entropy=0.83) demonstrated better fit across all measures w he n compared to a 2 class model. When the 3 class model was compared to a 4 class model, both BIC and ABIC (BIC 38235.764; ABIC=38137.250) indicated improvement in fit; the VLMR and BLRT tests also indicated that a 4 class model was significantly different compared to a 3 class model (p<0.01) even though the entropy value decreased to 0.77. The identified subtypes in the four class model were distinct in the patterns of substance use and demonstrated potential for interpretability. Further, shown in Table 5 3 is the quality of correct classification of youth into each of the latent classes which was fairly high The average conditional probability of being correctly classifi ed as a member of class 1 was 0.7 5 0.86 for class 2, 0.89 for class 3, and 0.74 for class 4. The 5 class model deteriorated appreciably in model fit from the 4 class model (BIC=38280.101; ABIC=38156.164; entropy 0.72). For these reasons, the 4 class solution which was comparatively parsimonious relative to other models and identified inte rpretable classes was favored over the other models.
92 The four classes were distinguished by distinct pattern s of past 30 day substance use (Figure 5 1 ). Class 1 was t h e largest class and comprised of youth with a high probability of low to no substance use (80.2%). Class 2 consisted primarily of users of alcohol with some marijuana use and a lower probability of us e of all other substances (13.0 %). Class 3 manifested a high probability of the use of alcohol, tobacco, marijuana and other illicit drugs with a low probability of non medical use of prescription drug s (4.7%). A high probability of non medical use of prescription drug s with use of other types of substances was observed with membership in Class 4 (2.1 %). We named these classes as low no substance use, alcohol marijuana use, alcohol, tobacco, marijuana and other illegal drug use (AT M+ ) and NMU of p rescription medication and other s ubstance use ( Rx++ ) subtypes respectively. Socio Demographic Characteristics b y LCA S ubtypes As shown in Ta ble 5 4 there were significant differences among the LCA defined subtypes when compared on socio demographic characteristics. Youth in the subtypes of alcohol marijuana ATM+ and Rx++ (mean: 16.23, 16.54 and 16.43 years respectively) were older than youth with low no substance use (mean: 14.81 years). Almost two thirds of youth in the Rx++ subtype was Caucasian while other subtypes had lower representation among Caucasians and ranged f rom 40.7 % ( low no substance use ) to 53.4% ( AT M+ ). Slightly higher representation of males was seen in AT M+ and Rx++ subtypes while there were a higher proportion of females in the low no substance use and alcohol marijuana use subtypes Compared to other subtypes, Rx++ followed by AT M+ were more likel y to have poorer grades, have sleep timings at or later than midnight, live in a single parent household and more likely to have fewer than 5 meals
93 per week with family. There were no differences in the area of residence by substance use subtype. Mental an d B ehavioral H ealth C haracteristics by LCA S ubtypes LCA defined subtypes significantly differed when compared on mental and behavioral health characteristics presented in Table 5 5 Overall, in a most interesting fashion, mental and behavioral health sympt oms increased with every level of increasing type and number of substances used; higher rates were observed among those with NMU of prescription medications and use of other substances ( Rx++ ) followed by alcohol, tobacco, marijuana and other drugs ( AT M+ ), with comparatively lower rates in alcohol marijuana users, and the least among youth with low no substance use Specifically, 69% of youth in Rx++ subtype r ated their health as fair to poor. E motional and mental health problems indicated by depression, anxiety symptoms, weight concerns and ADHD and CD symptoms were highest among youth in the ATM+ and Rx++ subtypes while rates were relatively lower for those in the alcohol marijuana subtype L ow no substance users were comparatively least likely to repor t emotional and mental health problems than the other subtypes Peer S ubstance U se by LCA S ubtypes The rates of peer substance use in the LCA identified subtypes are displayed in Table 5 5 which indicates that the highest rate is among youth with Rx++ at 85.9% whereas only 19.6 % of those in low no substance use classification reported having at least one friend with stimulant use. Multinomial L ogistic R egression A nalysis by LCA S ubtypes A latent class with covariates framework was used to determine the correlates of LCA identified subtypes using multinomial logistic regression analysis. LCA subtype
94 was the dependent variab le of interest with low no substance use as the referent group; the results are presented in Table 5 6 Relative to low no substances users, each year increase in age increased the odds for membership in the subtypes of alcohol marijuana use (AOR 1.35), AT M+ (AOR 1.3 7 ) and Rx++ subtypes (AOR 1.50) The socio demographic characteristics of Caucasian race, poor grades in school and sleep t imings at or after midnight increased the odds of membership into alcohol marijuana, AT M+ and Rx++ subtypes compared to the low no substance use subtype Positive family related characteristics of regular meals with family and living with both parents decr eased the odds of being in the subtypes of alcohol marijuana use, ATM+ and Rx++ use groups relative to the referent class of low no substance use Relative to urban areas, residing in rural areas decreased the odds of membership into alcohol marijuana use group Gender was not related to any LCA identified subtype of substance use relative to low no use subtype. A mong the mental and behavioral health problem indicators assessed, CD symptoms had the strongest association with substance use subtypes re lative to not reporting CD symptoms. CD symptoms increased the odds of AT M+ by over eight times (AOR 8. 68 ), almost six times for the Rx++ class (AOR 5 98 ) and about three times for alcohol marijuana use (AOR 2.9 6 ). The likelihood of ATM+ Rx++ and alcohol marijuana use increased with having weight concerns by 3 26 2 49 and 1.74 times compared to not having any weight concerns. D epressive and anxiety symptoms only slightly increased the odds of membership in all subtypes ( alcohol marijuana use, ATM+, Rx++ g roups ) relative to the low no use subtype Further, o nly Rx++ subtype was associated with having been diagnosed with ADHD (AOR 1. 63 ) Peer use of stimulant use was also
95 found to increase the likelihood of inclusion into all three subtypes of substance use ; membership in AT M+ increased by over 13 times, inclusion into the Rx++ subtype b y almost seven times and alcohol marijuana by three times when youth had at least one friend who used stimulants comp a red to not having a friend who used stimulant s Discus sion Our findings suggest heterogeneity in substance use patterns among a nationally representative sample of youth aged 10 to 18 years in the United States. More than a third of youth (37.2%) had used at least one type of substance in the past 30 days. Using LCA, we were able to empirically identify four subtypes of youth substance users: 1) non medical users of prescription stimulants, opioids and benzodiazepines and other substances (Rx++) ; 2) primarily alcohol, tobacco and other illicit drug users wit h low non medical use of prescription medication (AT M+ ); 3) primarily alcohol marijuana users ; and 4) low no substance users. In particular, o ur study found that a small but distinct group of youth (2%) engage d in the non medical use of prescription medica tion along with other substances ( Rx++ ) which distinguished these youth from the AT M+ group (4.7%) which used majority of other substances but had low non medical use of prescription medication These two subtypes were characterized by increased association with poor behavior al and mental health indicators, particularly CD symptoms and weight concerns. Rx++ and AT M+ users were also more likely to have a peer who used stimulants and had decreased parental monitoring indicators. The largest subtype consisted of low no use of substance (80.2%) and 13 % of youth were in the alcohol marijuana subtype; in comparison to the Rx++ and AT M+ subtypes, relatively few mental and behavioral health problems was observed among youth in these two group s. These findi ngs indicate diversity in youth substance use and find support in utilizing
96 person centered approaches to examine youth non medical use of prescription medication separately and together with other substance classes. Non medical Use of P rescription Stimul ants, Opioids Benzodiazepines and Other S ubstance U se ( Rx++ ) A subtype of youth (2 .1 %) demonstrated high likelihood of the non medical use of stimulants, opioids or benzodiazepines and the use of other illicit substances for youth (cigarette, alcohol, marijuana, other illicit drugs). Our study results indicated that non medical users of stimulants, opioids or benzodiazepines were included in the subtype which also used several other substance s ; the findings are similar to a study by Conway et al (2013) which used a nationally representative sample of 10 th grade students. However, a web based study conducted among high school adolescents in Michigan using LCA identified a subgroup of youth with relatively high rates of prescription drug mis use (characterized by both misuse and excessive medical use of prescription) and alcohol use with low rates of other substances (Cranford et al., 2013) The study also found female gender at increased risk of membership of this subtype of prescription drug misusers while our study did not find class membership to vary wit h gender. The differences in findings are most likely due to characteristics of the samples which greatly vary in terms of representativeness The Rx++ subtype in our study was characterized by strong as s ociations with CD symptoms Additionally, increas ed likelihood of non medical use of prescription medications and other substance use was linked to having ADHD symptoms, weight concerns, depression and anxiety symptoms and fair / poor self reported health. Their risk profile is further supported by signifi cant associations with having friends who used prescription stimulant medication. In particular, we found that CD was strongly
97 associated with Rx++ and AT M+ which is consistent with several prior studies (Costello 2007; Fergusson et al., 2007) It has been shown that while CD i ncrease s the risk of onset of all substance classes the risk associated with use of illicit substances is greater compared to licit substances (Hopfer et al., 2013) Further, i ncreased depressive and somatic symptoms among poly substance users were reported in a national sample of 10 th grade students (Conway et al., 2013) and among adolescent poly su bstance users attending urban primary health care clinics in Michigan (Bohnert et al., 2014) In particular increased association of depression as well as somatic symptoms characterized by headaches, stomachaches or backaches a mong students with m ultiple substance use have been found (Connell et al., 2010; Dierker et al., 2007) Our findings are in line with prior studies which have demonstrated tha t youth with use of multiple substance types have associated mental health needs which should be addressed along with efforts to reduce substance use (B ohnert et al., 2014; Conway et al., 2013) Additionally, socio demographic correlates for increased risk of Rx++ found in our study includes being older, of Caucasian race and having poor grades which confirm findings reported in prior studies (Bohnert et al., 2014) Family related factors such as living in single parent household and not eating meals regularly with family were positively associated with the risk of inclusion in the Rx++ subtype These findings may reflect poorer levels of engagement or cohesion in the family which has been shown to exert detrimental influences in subs tance use (Taylor et al., 2012) Further, in light of the non medical use of prescription medication occurring largely via friend s parental or known sources (Cottler et al., 2013; Johnston et al., 2014; Schepis and
98 Kris hnan Sarin, 2008) parental monitoring should be recommended which can be helpful for families with high risk profiles for substance use. Alcohol, T obacco, M arijuana and O ther I llicit D rug U sers ( ATM+ ) Our study identified a small group of youth who were predominantly AT M+ While this group demonstrated almost similar rates of AT M+ compared to the Rx++ subtype this group was distinct in their low use of prescription medication. Although this subtype was characterized by similar risk factors as the Rx++ subtype the rates and associated odds ratios were comparatively lower. However, CD symptoms showed the strongest association with AT M+ increasing the risk by over eight times compared to those without CD The presence of CD has been associated with a particularly elevated risk for substance use disorders and also with the persistence of other mental disorders such as depression and anxiety (Nock et al., 2006) Socio demographic correlates for increased risk of AT M+ included older age, poor grades, living in a single parent household, less frequent meals with family in a week and sleep timings later than midnight. These you th were also likely to have at least one friend who used prescription stimulants. Poor grades were positively associated with both substance use subtypes, a finding that is congruent with a large body of research describing the association between lower ac ademic achievement and substance use (Bachman et al., 2011; Bohnert et al., 2014; Wheeler, 2010) Alcohol and M arijuana U ser s A third subtype of youth who were primarily alcohol users with some use of marijuana and low rates of all other types of substances was identified in our study This finding is consistent with the study by Conway et al (2013) us ing baseline data of a national longitudinal study of 10 th graders that found a subtype that consisted of
99 predominantly alcohol users and another subtype that comprised of primarily marijuana users only. A number of prior studies have also indicated a subtype with high probability of alcohol and tobacco use or use of tobacco, alcohol and marijuana (Cleveland et al., 2010; Connell et al., 2010) Tobacco users in our study find representation in the Rx++ and the AT M+ subtypes but not in the predominantly alcohol using subgroup of youth; this finding may reflect the decreasing trend of tobacco use rates which has been documented in both national and regional studies (Gilreath et al., 2014) The alcohol marijuana user subtype in our study was characterized by older age, Caucasian race, poor grades, late bed timings, wei ght concerns, CD symptoms depression and anxiety symptoms and peer stimulant use. Youth living in rural areas had decreased risk of membership in this subtype relative to those in urban areas. Low no S ubstance U sers A significant proportion of youth consi sted of a subtype with comparatively low or no substance use, a finding consistent with other national studies that focus on youth substance use behavior (Conway et al., 2013) Their reduced risk profile was supported by reduced odds of behavioral and mental health issues, and peer stimulant use as compared with other substance use subtypes. Although this group demonstrated lesser involvement with all substance use types, some alcohol use was observed directing attention to what maybe the normative behavior among youth in contemporary times. Other studies carried out on youth in other populations such as primary care clinics and students have also established that the large majority of youth abstain from substance use (Bohnert et al., 2014; Cranford et al., 2013; Gilreath et al., 2014)
100 Study L imitations and S trengths The findings of this study should be interpreted with the consideration of s ome limitations. The N MAPSS used self reports to assess for past 30 day substance use as well as the risk factors, which may be prone to under reporting owing to social desirability and recall errors. The cross sectional nature of the study only allows th e estimation of associations between substance use subtypes without the abil ity to make causal inferences. Recruitment of youth was not carried out using random sampling techniques and some youth who do not frequent entertainment venues might have been mis sed. Further, the period of substance use was not permissible with the present data which would have been more informative in terms of distinguishing youth based on persistent patterns of substance use. The N MAPSS data has noteworthy strengths. The large size and recruitment of the sample from youth friendly venues allowed the inclusion of school drop outs and home schooled youth in the study sample. Further, the sample is highly representative of the US population aged 10 to 18 years suggesting the high generalizability of study findings (Cottler et al, 2013) The study is also one of very few to assess substance use in youth younger than 12 years of age. The detailed assessment in the N MAPPS study has allowed the characterization of contemporary patter ns of substance use and abuse and its associated risk factors among youth using a person centered approach. Study findings indicate that substance use continues to be a public health problem among youth. Further substance use among youth is linked to unmet mental health needs It is known that the majority of youth in need of mental health services do not receive services I mproving access to
101 mental health services and care may play an important role in the prevention of substance abuse and dependence in yo uth.
102 Table 5 1. Substance use among youth in the N MAPSS (n=11 007) Substance Use % Cigarette use 9.9 % Alcohol use 27.0 % Marijuana use 17.2 % Non medical stimulant use 3.6 % Non medical opioid use 2.9 % Non medical benzodiazepine use 2.1 % Other illicit drug use 11.7 % At least one substance use 37.2% Table 5 2. Statistical fit measures for substance use subtypes of youth in the N MAPSS (n=11 007) Model fit statistics Class 1 Class 2 Class 3 Class 4 Class 5 AIC 46338.204 38596.377 38084.425 38009.269 37995.156 BIC 46389.348 38705.971 38252.469 38235.764 38280.101 AIC 46367.103 38658.303 38179.378 38137.250 38156.164 VLMR 0.0000 0.0000 0.0016 0.0061 BLRT 0.0000 0.0000 0.0017 0.0065 Entropy 0.849 0.829 0.767 0.724 Table 5 3 Average latent class probabilities for most likely latent class membership for substance use (row) by latent class (column) youth in the N MAPSS. Latent class Low no use (n 8827) Alcohol marijuana (n 1435) AT M+ (n 517) Rx++ (n 228) Class 1 0.748 0.100 0.000 0.151 Class 2 0.088 0.860 0.004 0.048 Class 3 0.002 0.0000 0.892 0.106 Class 4 0.176 0.013 0.070 0.741 Table 5 4 Demographic characteristics of substance use subtypes of youth in the N MAPSS. Characteristic Low no use (n 8827) Alcohol marijuana (n 1435) ATM+ (n 517) Rx++ (n 228) p value Age (mean) 14.8 16.2 16.5 16 .4 <.0001*** Gender 46. 6% 51.1 % 55.9 % 55.7 % <.0001*** Caucasian 40.7 % 49.5 % 53.4 % 63.9 % <.0001*** Residence Urban 47.8 % 47.9 % 45.4 % 40 .3 % Suburban 37.1 % 37.5 % 40.2 % 38.2 % 0.0875 Rural 15.1 % 14.6 % 14.3 % 21.5 % Poor grades ( D & F 20.0 % 34.4 % 51.8 % 47.4 % <.0001*** Meals ( 5 per week) 45.6 % 25.8 % 18.6 % 17.2 % <.0001*** Sleep timings (12am or later) 18.5 % 34.9 % 4 5.6 % 49.1 % <.0001*** Dual parent household 58.3 % 47.1 % 41.6 % 37.3 % <.0001*** p <.05, ** p <.001, *** p <.0001 ; Note: ADHD: Attention Deficit Hyperactivity Disorder; CD : Conduct Disorder
103 Table 5 5 Health characteristics of substance use subtypes of youth 10 to 18 years of age in the N MAPSS. Characteristic Low no use (n 8827) Alcohol marijuana (n 1435) ATM+ (n 517) Rx++ (n 228) p value Health rating ( fair/poor ) 4.7 % 9.5 % 16 5 % 69. 0 % <.0001 *** ADHD diagnosis 9.8 % 18.5 % 33.7 % 31.6 % <.0001 *** CD symptoms ( 3) 7.2 % 27.1 % 49.9 % 58.3 % <.0001 *** Depressi ve symptoms 47.8 % 64.5 % 73.6 % 78.5 % <.0001 *** Anxiety symptoms 22.4 % 35.8 % 44.5 % 53.1 % <.0001 *** 2.7 % 7.4 % 12.4 % 18.0 % <.0001 *** Friend stimulant use 19.6 % 52.4% 72.7 % 85.9 % <.0001 *** p <.05, ** p <.001, *** p <.0001 Note: ADHD Attention Deficit Hyperactivity Disorder; CD Conduct Disorder Table 5 6 Correlates of substance use subtypes of youth in the N MAPSS (n=11,007) Characteristic AOR 95% CI Age (years) Low no use ref Alcohol marijuana 1.35 (1.30 1.40) A TM+ 1.37 (1.24 1.51) Rx++ 1.50 (1.40 1.60) Male gender Low no use ref Alcohol marijuana 0.99 (0.87 1.13) A TM+ 0.99 (0.72 1.38) Rx++ 1.03 (0.83 1.29) Caucasian race Low no use ref Alcohol marijuana 1.37 (1.20 1.56) A TM+ 1.95 (1.42 2.68) Rx++ 1.43 (1.15 1.78) Residence Low no use ref Alcohol marijuana (Suburban vs. Urban) 0.96 (0.84 1.10) Alcohol marijuana (Rural vs. Urban) 0.82 (0.68 0.99) A TM+ (Suburban vs. Urban) 1.04 (0.74 1.45) A TM+ (Rural vs. Urban) 1.15 (0.55 1.04) Poly sub (Suburban vs. Urban) 1.04 (0.83 1.30) Poly sub (Rural vs. Urban) 0.76 (0.55 1.04) 4 meals or more with family a week Low no use ref Alcohol marijuana 1.69 (1.31 2.17) A TM+ 1.75 (1.19 2.57) Rx++ 1.49 (1.29 1.70) Low no use ref Alcohol marijuana 1.48 (1.29 1.71) A TM+ 1.77 (1.29 2.43) Rx++ 2.34 (1.88 2.91)
104 Table 5 6 continued. Characteristic AOR 95% CI Lives in single parent household Low no use ref Alcohol marijuana 1.26 (1.11 1.43) A TM+ 1.60 (1.17 2.18) Rx++ 1.29 (1.04 1.59) Sleep after 12 pm Low no use ref Alcohol marijuana 1.39 (1.21 1.60) A TM+ 1.97 (1.45 2.69) Rx++ 1.63 (1.32 2.02) ADHD Low no use ref Alcohol marijuana 1.09 (0.91 1.30) A TM+ 1.18 (0.84 1.65) Rx++ 1.63 (1.28 2.06) CD symptoms Low no use ref Alcohol marijuana 2.96 (2.51 3.49) A TM+ 8.68 (6.23 12.10) Rx++ 5.98 (4.74 7.53) Depressive symptoms Low no use ref Alcohol marijuana 1.32 (1.14 1.53) A TM+ 1.63 (1.19 2.24) Rx++ 1.71 (1.37 2.13) Anxiety symptoms Low no use ref Alcohol marijuana 1.19 (1.04 1.37) A TM+ 1.82 (1.338 2.48) Rx++ 1.35 (1.09 1.68) Weight concerns Low no use ref Alcohol marijuana 1.74 (1.34 2.27) A TM+ 3.26 (2.08 5.10) Rx++ 2.49 (1.75 3.56) Friend stimulant use Low no use ref Alcohol marijuana 3.17 (2.79 3.61) A TM+ 13.61 (9.15 20.48) Rx++ 6.59 (5.26 8.26)
105 Figure 5 1 Item probability for latent class subtypes of youth in the four class m odel (N=1 1 ,007) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Tobacco Alcohol Marijuana Other illicit drugs Stimulants Opioids Benzodiazepines Rx++ (2.1%) ATM+ (6.2%) Alcohol-marijuana (18.9%) No-low Substance Use (80.2%)
106 CHAPTER 6 CONCLUSION The widespread use of pharma cotherapy for the treatment of A ttention Deficit Hyperactivity Disorder has greatly increased the environmental availability of prescription stimulants in the United States. I ncreased prescribing and production of ADHD stimulants has been accompanied by reports of the non medical use of stimulants prescription in the last two decade s Youth and young adults have been indicat ed to be at highest risk for the non medical use of stimulants ; about 5 % of those aged 10 to 18 years engage in past 30 day non medical use of stimulants Of concern are the increased r eport s of adverse physical complications of non medical use that is reflected by a significant rise in emergency department visits that involve stimulant misuse. So me efforts have been made to examine epidemiological characteristics of non medical use of stimulants among youth using both national and regional samples. However, study results are not comparable because of the use of variable definitions in the measurem ent of non medical use and because all stimulants commonly prescribed have not been assessed even by the existing national studies. Further, in the efforts to understand the non medical use of stimulants among youth increased attention has been given to a few factors such as socio demographic or perceived risks and harm related to stimulants while others such as mental and behavioral health considered important in other substance use literature have not been given its due focus. It is known that substance use and misuse and mental health problems co occur, and each can influence the course and outcome of the other Additionally, youth may misuse more than one substance type; polysubstance use among youth with non medical use of stimulants has been commonly reported. Substance use and m ental health problems
107 among youth relate to increased social and economic costs, as they often develop into more disabling conditions later in life associated with poorer outcomes Preventive efforts are emphasized to help forestall the development and progression of substance misuse and mental disorders, and early intervention can limit their severity. One of the most urgent tasks in addressing youth non medical use of prescription stimulants is improving and expanding the evidence base using samples that are representative to ensure generalizability of study findings Systematic research on the prevalence nature and determinants of non medical use of stimulants among youth and on the prevention, early intervention and treatment strategies will be critical to ensuring youth health a nd development Thus, i n order to improve understanding of the magnitude and mental and behavioral risk factors for the public health problem of non medical use of stimulants the present stu dy assessed the correlates of past 30 day non medical use of stimulants in terms of externalizing and internalizing problem behaviors using a representative national sample of youth aged 10 to 18 years in the US We found 3.6 % of youth aged 10 to 18 years reported the non medical use of prescription stimulants. Overall, our sample showed high rates of externalizing problems of ADHD and CD problem s with higher rates among those who used stimulants non medically. Rates of internalizing problems of depressive and anxiety symptoms and weight concerns were comparable to other studies, a lthough rates were higher in those with non medical use of stimulants. Multinomial logistic regression models used indicated that t he strongest association was between peer influence measured as having at least one friend who used a stimulant which increased the likelihood of non medical use of stimulants by about six times E xternalizing and internalizing problems
108 also increased the ris k of non medical use of stimulants in youth. ADHD and CD symptoms independently increased the likelihood of non medical use of stimulants as did the use of other substances of abuse such as cigarettes, alcohol, marijuana and other illicit drugs. Internaliz ing problems of depressive symptoms and weight concerns were associated with increased risk of non medical use of stimulants We did not find gender to moderate the association between externalizing problem behaviors of ADHD or CD, and the non medical use of prescription stimulants With a view to aid strategic and targeted prevention efforts, we also aimed to assess the critical ages when youth are most likely to initiate non medical use of stimulants and inform the public health community regarding factor s that are associated with the onset. Using survival analysis we modeled the time to onset of non medical use of stimulants and found that prior to age 12 years, the rate of non medical use of stimulants was low and below 2%. However, with each year incre ase in age at 13 years, the risk of initiating non medical use of stimulants increased and peak ed at 18 years of age with an incidence of 6.9%. We used Cox regression models to assess factors that were associated with the hazards of initiating non medical use of stimulants Being male, Caucasian and having ADHD symptoms increased the hazards of initiating non medical use of stimulants. Further, the u se of substances such as alcohol and marijuana prior to the non medical use of stimulants w as also associated with earlier non medical use of stimulants compared to those who used later. There have been reports of increased rates of the use of other substance s among youth with non medical use of stimulants a s well as co ingestion of stimulants with substances such as marijuana, alcohol and other illicit drugs We were interested to
109 contextualize non medical use of stimulants within the larger substance use and misuse public health problem among youth. Thus, we aimed to empirically identify subtypes o f youth based on their patterns of use of seven substance classes ( cigarette s alcohol, marijuana, other illicit drugs, and non medical use of stimulants, opioids and benzodiazepines ) Additionally, we examined whether youths mental and behavioral health characteristics varied by the substance use subtypes. Overall, 37% of youth had used at least one of the seven types of substance in the past 30 days. L atent class analysis identified four subtypes of substance use patterns : 1) strong representation of non medical use of prescription medication (stimulants, sedatives and benzodiazepines) in a small but distinct group of youth (2.1% ; Rx++ ) ; 2) p rimarily alcohol, tobacco and other illicit drug users with low non medical use of prescription medication ( 4.7%; AT M+ ) group that represented about 5% of youth ; 3) primarily alcohol marijuana users consisting of 13% of youth and 4) low no substance users that was the largest group (80%) We also examined the mental and behavioral health characteristics of the identif ied subtypes of substance use. The Rx++ and AT M+ subtypes were characterized by an increased association with behavioral and mental health problems and decreased parental monitoring indicators compared to the other two subtypes alcohol and marijuana use g roup and the low no use group. Overall, t he present study has identified a number of risk factors associated with the non medical use of stimulants. As noted, an important risk factor is peer stimulant use that seems to be highly associated. The study fin dings suggest that non medical use of stimulants is not an isolated behavior and seem to be part of a broader group of behaviors that are linked. In particular externalizing behaviors of ADHD and other
110 substance use are important factors associated with bo th the onset and past 30 day non medical use of stimulants among youth It is also clear that youth who engage in non medical use of stimulants are also more likely to report the use of other drugs. Multiple studies including the present study have found associations between non medical use of stimulants and increased rates of cigarette, alcohol, marijuana, and other illicit drug use among youth in the United States. There seems to be indications of internalizing problems such as depressive symptoms and we ight concerns that may be linked to the non medical use of prescription stimulants. Thus, study findings suggest that targeting mental health problems among youth such as ADHD symptoms and preventing substance use may prevent non medical use of stimulants in a substantial proportion of youth. It is estimated that 1 in 6 youth are exposed to prescription stimulants either medically or non medically. Further, it has perception of the risk of non medical use of medication in general is low. A large majority may not be aware of their problematic non medical use and a t this time, there are no FDA approved medications for treating stimulant addiction. The high level of exposure to stimulants among youth and the unavailability of pharmac otherapy to treat stimulant addiction highlights the importance of increasing prevention efforts before non medical use develops into dependence or risk of adverse health outcomes increase Prevention activities should ideally begin before the teen years w hen the risk of initiation begins to first set in. An important aspect that emerged is one of improving access to treatment for the modifiable risk factors identified by the present st udy, such as Attention Deficit H yperactivity Disorder and other substan ce use that are associated or co occur with the
111 non medical use of prescription stimulants. Additionally, physicians and pharmacists can play an important role in the identification and prevention of prescription stimulant abuse. U p to date knowledge on ap propriate prescribing practices for stimulants particularly among youth with ADHD should be regularly provided. Youth and their parents (who visit clinics or the pharmacy) can be educated by p hysician s and pharmacists on the responsible and appropriate man ner of use and storage of medications. P hysicians can be more alert when prescribing or identifying misuse among those who are prescribed the medication B ecause a large majority of non medical users of stimulants obtain stimulant medication from people co mmonly known to them such as family or friends, physicians and pharmacists can also caution youth against the diversion of stimulants and the associated adverse effects either duri ng prescribing or dispensing of stimulant m edications Further, a worrisome misconception among youth is that the medications are safer to abuse than other illicit drugs because they are prescribed by medical professionals. Another common belief among youth is that stimulants improve cognitive ally have ADHD which has been shown to be untrue and in fact is related to poorer grades. School and college substance abuse prevention efforts should work on clarifying these erroneous beliefs and stress on the dangers of non medical use. Non medical use of prescription medication is a complex problem which requires a multi pronged, strategic and sustained effort which can be achieved only with coordinated efforts at all levels that range from public health to clinical medicine, public safety, and other stakeholders.
112 Lastly, the present study and the majority of studies have focused on the non medical use of stimulants without focusing on whether youth use these medications persistently. Future efforts can be directed towards characterizing youth who non medically use occasionally and those who are more regular users. Efforts to examine duration of use of stimulants among youth should be carried out. Further research is needed to identify those individuals whose non medical use develops into dependence lat er in life. Intervention programs for youth with non medical use of stimulants must be developed, and the outcomes evaluated.
113 APPENDIX A NATIONAL MONITORING OF PRESCRIPTION STIMULANT S SURVEY
136 APPENDIX B SENSITIVITY ANALYSIS Table B 1 Estimated hazards of lifetime prescription stimulant use initiation among youth 10 to 18 years of age in the N MAPPS Characteristic Model 1 a AHR (95% CI) Model 2 b AHR (95% CI) Model 3 c AHR (95% CI) Model 4 d AHR (95% CI) Gender Female r ef r ef r ef r ef Male 1.11 (0.92 1.34) 1.08 (0.91 1.26) 1.09 (0.93 1.27) 1.08 (0.89 1.30) ADHD symptoms No r ef r ef r ef r ef Yes 3.82 (3.00 4.85) 1.43 (1.20 1.70) 4.13 (3.46 4.92)*** 4.64 (3.75 5.74) a Final model with MU only and both excluded; b Model with MU only coded 0 and both coded as NMU; c Model with MU excluded and both coded as NMU; d Model with MU coded as non use and both excluded; Note: ADHD Attentio n Deficit Hyperactivity Disorder
137 LIST OF REFERENCES analytic study. Psychol. Monogr. 80, 1 37. Achenbach, T.M., Edelbrock, C.S., 1978. The classification of child psychopathology: a revie w and analysis of empirical efforts. Psychol. Bull. 85, 1275 1301. Ali, M.M., Dean, D., Lipari, R., Dowd, W.N., Aldridge, A.P., Novak, S.P., 2015. The mental health consequences of nonmedical prescription drug use among adolescents. J. Ment. Health Policy Econ. 18, 3 15. Arcos Burgos, M., Vlez, J.I., Solomon, B.D., Muenke, M., 2012. A common genetic network underlies substance use disorders and disruptive or externalizing disorders. Hum. Genet. 131, 917 929. Arria, A.M., Garnier Dykstra, L.M., Caldeira, K E.D., 2011. Persistent nonmedical use of prescription stimulants among college students: Possible association with ADHD symptoms. J. Atten. Disord. 15, 347 356. Arria, A.M., Mericle, A.A., Meyers, K., Winters, K.C. 2012. Parental substance use impairment, parenting and substance use disorder risk J. Subst. Abuse Treat. 43, 114 122. Austic, E., McCabe, S.E., Stoddard, S.A., Ngo, Q.E., Boyd, C., 2015 Age and cohort patterns of medical and nonmedical use of control led medication among adolescents J. Addict. Med. 9, 376. Austic, E., 2015. Peak ages of risk for starting nonmedical use of prescription stimulants. Drug Alcohol Depend. 152, 224 229. ace, J.M., 2011. Racial/ ethnic differences in the relationship between parental education and substance use among U.S. 8th 10th and 12th grade students : Findings from the Monitoring the Future Project. J. Stud. Alcohol Drugs 72, 279 285. Back, S.E., P ayne, R.L., Simpson, A.N., Brady, K.T., 2010. Gender and prescription opioids: findings from the National Survey on Drug Use and Health. Addict. Behav. 35, 1001 1007. Bali, V., Raisch, D.W., Moffett, M.L., Khan, N., 2013. Determinants of nonmedical use, a buse or dependence on prescription drugs, and use of substance abuse treatment. Res. Soc. Adm. Pharm. 9, 276 287. Barkley, R.A., Fischer, M., Smallish, L., Fletcher, K., 2004. Young adult follow up of hyperactive children: antisocial activities and drug u se. J. Child Psychol. Psychiatry 45, 195 211.
138 Baron, R.M., Kenny, D.A., 1986. The moderator mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 51, 1173 1182. Baum eister, S.E., Tossmann, P., 2005. Association between early onset of cigarette, alcohol and cannabis use and later drug use patterns: an analysis of a survey in European metropolises. Eur. Addict. Res. 11, 92 98. Behrendt, S., Wittchen, H. U., Hfler, M., Lieb, R., Beesdo, K., 2009. Transitions from first substance use to substance use disorders in adolescence: Is early onset associated with a rapid escalation? Drug Alcohol Depend. 99, 68 78. Benotsch, E.G., Koester, S., Luckman, D., Martin, A.M., Cejka, A., 2011. Non medical use of prescription drugs and sexual risk behavior in young adults. Addict. Behav. 36, 152 155. Berenson, A.B., Rahman, M., 2011. Prevalence and correlates of prescription drug misuse among young, low income women receiving public he althcare. J. Addict. Dis. 30, 203 215. Bernstein, 2015. Conduct disorder: Overview, frequency, clinical course. Available at: http://emedicine.medscape.com/article/918213 overview Blanco, C., Alderson, D., Ogburn, E., Grant, B.F., Nunes, E.V., Hatzenbuehler, M.L., Hasin, D.S., 2007. Changes in the prevalence of non medical prescription drug use and drug use disorders in the United States: 1991 1992 and 2001 2002. Drug Alcohol Depend 90, 252 260. Bloom, B., Jones, L.I., Freeman, G., 2013. Summary health statistics for U.S. Children: National Health Interview Survey 2012 (No. 10(258)), Vital Health Stat. National Center for Health Statistics. Bogle, K.E., Smith, B.H., 2009. Illicit Methylphenidate Use: A review of prevalence, availability, pharmacology, and consequences Curr. Drug Abuse Rev. 2, 157 176. Bohnert, K.M., Walton, M.A., Resko, S., Barry, K.T., Chermack, S.T., Zucker, R.A., Zimmerma n, M.A., Booth, B.M., Blow, F.C., 2014. Latent class analysis of substance use among adolescents presenting to urban primary care clinics Am. J. Drug Alcohol Abuse 40, 44 50. Boyd, C.J., McCabe, S.E., Cranford, J.A., Young, A., 2007. Prescription drug a buse and diversion among adolescents in a southeast Michigan school district. Arch. Pediatr. Adolesc. Med. 161, 276 281. motivations to abuse prescription medications Pediatrics 118 2472 2480.
139 nonmedical use of prescription medications and other problem behaviors J. Adolesc. Health Off. Publ. S oc. Adolesc. Med. 45, 543 550. Bracken, B.K., Rodolico, J., Hill, K.P., 2013 Sex, age, and progression of drug use in adolescents admitted for substance use disorder treatment in the northeastern United States: Comparison with a national survey. Subst. Abuse Off. Publ. Assoc. Med. Educ. Res. Subst. Abuse 34, 263 272. Bright, G.M., 2008. Abuse of medications employed for the treatment of ADHD: Results from a large scale community survey Medscape J. Med. 10, 111. Brinkman, W.B., Epstein, J.N., Auinger, P., Tamm, L., Froehlich, T.E., 2015. Association of Attention Deficit/Hyper activity Disorder and Conduct Disorder with early tobacco and alcohol use. Drug Alcohol Depend. 147, 183 189. Brook, J.S., Balka, E.B., Zhang, C., Brook, D.W., 2014. ADHD, conduct disorder, substance use disorder, and nonprescription stimulant use J. Att en. Disord. 1087054714528535. Brook, J.S., Duan, T., Brook, D.W., 2007. Fathers who abuse drugs and their adolescent children : Longitudinal predictors of adolescent aggression Am. J. Addict. 16, 410 417. Butterworth, P., Slade, T., Degenhardt, L., 2014. Factors associated with the timing and onset of cannabis use and cannabis use disorder: Results from the 2007 Australian National Survey of Mental Health and Well Being. Drug Alcohol Rev. 33, 555 564. Buu, A., DiPiazza, C., Wang, J., Puttler, L.I., Fitzge rald, H.E., Zucker, R.A., 2009. Parent, family, and neighborhood effects on the development of child substance use and other psychopathology from preschool to the start of adulthood J. Stud. Alcohol Drugs 70, 489 498. Campbell, S.B., Shaw, D.S., Gilliom, M., 2000. Early externalizing behavior problems: toddlers and preschoolers at risk for later maladjustment. Dev. Psychopathol. 12, 467 488. Casey, B.J., Jones, R.M., 2010. Neurobiology of the adolescent brain and behavior: implications for substance use d isorders J. Am. Acad. Child Adolesc. Psychiatry 49, 1189 1201. Cassidy, T.A., McNaughton, E.C., Varughese, S., Russo, L., Zulueta, M., Butler, S.F., 2013. Nonmedical use of prescription adhd stimulant medications among adults in a substance abuse treatm ent population : Early findings from t he NAVIPPRO Surveillance System. J. Atten. Disord. 1087054713493321.
140 Castle, L., Aubert, R.E., Verbrugge, R.R., Khalid, M., Epstein, R.S., 2007. Trends in medication treatment for ADHD. J. Atten. Disord. 10, 335 342. Catalano, R.F., White, H.R., Fleming, C.B., Haggerty, K.P., 2011. Is nonmedical prescription opiate use a unique form of illicit drug use ? Addict. Behav. 36, 79 86. Centers for Disease Control and Prevention, 2014. 1991 2013 High School Youth Risk Behavio r Survey Data. Centers for Disease Control and Prevention (CDC). Challman, T.D., Lipsky, J.J., 2000. Methylphenidate: its pharmacology and uses. Mayo Clin. Proc. 75, 711 721. Charach, A., Yeung, E., Climans, T., Lillie, E., 2011. Childhood Attention Defic it/Hyperactivity Disorder and future substance use disorders : Comparative meta analyses J. Am. Acad. Child Adolesc. Psychiatry 50, 9 21. Chen, L. Y., Strain, E.C., Alexandre, P.K., Alexander, G.C., Mojtabai, R., Martins, S.S., 2014a. Correlates of nonmed ical use of stimulants and methamphetamine use in a national sample. Addict. Behav. 39, 829 836. Chen, X., Unger, J.B., 1999. Hazards of smoking initiation among Asian American and non Asian adolescents in California: A Survival Model Analysis. Prev. Med. 28, 589 599. middle and high school students in Tennessee. J. Drug Issues 38, 467 492. Chung, T., Martin, C.S., 2001. Classification and course of alcohol problems among ado lescents in addictions treatment programs Alcohol. Clin. Exp. Res. 25, 1734 1742. Clark, T.T., Doyle, O., Clincy, A., 2013. Age of first cigarette, alcohol, and marijuana use among U.S. biracial/ethnic youth : A population based study Addict. Behav. 38, 2450 2454. Clemow, D.B., Walker, D.J., 2014. The potential for misuse and abuse of medications in ADHD: a review. Postgrad. Med. 126, 64 81. Cleveland, M.J., Collins, L.M., Lanza, S.T., Greenb erg, M.T., Feinberg, M.E., 2010 Does individual risk moderate the effect of contextual level protective factors ? A Latent class analysis of substance use J. Prev. Interv. Community 38, 213 228. Collins, D., Abadi, M.H., Johnson, K., Shamblen, S., Thompson, K., 2011. Non medical use of prescription drugs among yout h in an Appalachian population: prevalence, predictors, and implications for prevention. J. Drug Educ. 41, 309 326.
141 Conn, B.M., Marks, A.K., 2014. Ethnic/Racial differences in peer and parent influence on adolescent prescription drug misuse. J. Dev. Behav. Pediatr. 35, 257 265. Connell, C.M., Gilreath, T.D., Aklin, W.M., Brex, R.A., 2010. Social ecological influences on patterns of substance use among non metropolitan high school students. Am. J. Community Psychol. 45, 36 48. Conway, K.P., Vullo, G.C., Nichter, B., Wang, J., Compton, W.M., Iannotti, R.J., Simons Morton, B., 2013. Prevalence and P atterns of polysubstance use in a nationally representative sample of 10th graders in the United States. J. Adolesc. Health Off. Publ. Soc. Adolesc. Med. 52, 716 723. Costello, E.J., 2007. Psychiatric predictors of adolescent and young adult drug use and abuse: What have we learned? Drug Alcohol Depend., Psychiatric Predictors of Early Drug Use and Abuse 88, Supplement 1, S97 S99. Cottler, L.B., Striley, C.W., Lasopa, S.O., 2013. Assessing prescription stimulant use, misuse, and diversion among youth 10 18 years of age: Curr. Opin. Psychiatry. 26, 511 519. Cotto, J.H., Davis, E., Dowling, G.J., Elcano, J.C., Staton, A.B., Weiss, S.R.B., 2010. Gender effects on drug use, abuse, and dependence: A special analysis of results from the national survey on drug use and health. Gend. Med. 7, 402 413. nonmedical use and excessive medical use of prescription medications and the identification of substance use subgroups Addict. Behav. 38. Daniulaityte, R., Falck, R., Wang, J., Carlson, R., 2009. Illicit Use of Pharmaceutical Opioids Among Young Polydrug Users in Ohio. Addict. Behav. 34, 649 653. Darke, S., K aye, S., Torok, M., 2012. Age related patterns of drug use initiation among polydrug using regular psychostimulant users. Drug Alcohol Rev. 31, 784 789. Davis, C.E., Hyde, J.E., Bangdiwala, S.I., Nelson, J.J., 1986. An example of dependencies among variab les in a conditional logistic regression, in: Modern Statistical Methods in Chronic Disease Epidemiology. 140 147. Degenhardt, L., Chiu, W.T., Conway, K., Dierker, L., Glantz, M., Kalaydjian, A., Merikangas, K., Sampson, N., Swendsen, J., Kessler, R.C., 20 09. Does the development of drug dependence in the National Comorbidity Study Replication. Psychol. Med. 39, 157 167. DeLisi, M., Angton, A., Behnken, M.P., Kusow, A.M., 2015. Do adolescent drug users fare the worst? onset type, juvenile delinquency, and criminal careers Int. J. Offender Ther. Comp. Criminol. 59, 180 195.
142 Dierker, L.C., Vesel, F., Sledjeski, E.M., Costello, D., Perrine, N., 2007. Testing the dual pathway hypot hesis to substance use in adolescence and young adulthood. Drug Alcohol Depend. 87, 83 93. Dom, G., Hulstijn, W., Sabbe, B., 2006. Differences in impulsivity and sensation seeking between early and late onset alcoholics. Addict. Behav. 31, 298 308. Duka T., 2011. Deficits in reward system activation and impulsivity: a predisposing condition to alcoholism? Biol. Psychiatry 69, 612 613. Dussault, C.L., Weyandt, L.L., 2013. An examination of prescription stimulant misuse and psychological variables among s orority and fraternity college populations J. Atten. Disord. 17, 87 97. Eisenberg, N., Cumberland, A., Spinrad, T.L., Fabes, R.A., Shepard, S.A., Reiser, M., Murphy, B.C., Losoya, S.H., Guthrie, I.K., 2001. The relations of regulation and emotionality to Child Dev. 72, 1112 1134. Fergusson, D.M., Horwood, L.J., Ridder, E.M., 2007. Conduct and attentional problems in childhood and adolescence and later substance use, abuse and dependence: result s of a 25 year longitudinal study. Drug Alcohol Depend. 88, S14 26. Fischer, B., Bibby, M., Bouchard, M., 2010. The Global Diversion of Pharmaceutical Drugs Non medical use and diversion of psychotropic prescription drugs in North America: a review of sou rcing routes and control measures. Addiction 105, 2062 2070. Fleary, S.A., Heffer, R.W., McKyer, E.L.J., 2011. Dispositional, ecological and biological influences on adolescent tranquilizer, Ritalin, and narcotics misuse. J. Adolesc. 34, 653 663. Flory, K., Milich, R., Lynam, D.R., Leukefeld, C., Clayton, R., 2003. Relation between childhood disruptive behavior disorders and substance use and dependence symptoms in young adulthood: individuals with symptoms of attention deficit/hyperactivity disorder and conduct disorder are uniquely at risk. Psychol. Addict. Behav. J. Soc. Psychol. Addict. Behav. 17, 151 158. Ford, J.A., 2009. Misuse of over the counter cough or cold medications among adolescents : Prevalence and correlates in a national sample J. Adolesc Health 44, 505 507. Ford, J.A., Lacerenza, C., 2011. The relationship between source of diversion and prescription drug misuse, abuse, and dependence Subst. Use Misuse 46, 819 827. Ford, J.A., McCutcheon, J., 2012. The misuse of Ambien among adolesc ents: Prevalence and correlates in a national sample. Addict. Behav. 37, 1389 1394.
143 Ford, J.A., Reckdenwald, A., Marquardt, B., 2014. Prescription drug misuse and gender Subst. Use Misuse 49, 842 851. Frodl, T., 2010. Comorbidity of ADHD and Substance Use Disorder (SUD): A Neuroimaging Perspective. J. Atten. Disord. 14, 109 120. Galanti, M.R., Rosendahl, I., Post, A., Gilljam, H., 2001. Early gender differences in adolescent tobacco use The experience of a Swedish cohort. Scand. J. Public Health 29, 314 317. Garfield, C.F., Dorsey, E.R., Zhu, S., Huskamp, H.A., Conti, R., Dusetzina, S.B., Higashi, A., Perrin, J.M., Kornfield, R., Alexander, G.C., 2012. Trends in Attention Deficit Hyperactivity Disorder (ADHD) ambulatory diagnosis and medical treatment in the United States, 2000 2010. Acad. Pediatr. 12, 110 116. Garnier Nonmedical use of prescription stimulants during college: Four year trends in exposure opportunity, use, motives, and so urces. J. Am. Coll. Health. 60, 226 234. Gil, A.G., Wagner, E.F., Tubman, J.G., 2004. Associations between early adolescent substance use and subsequent young adult substance use disorders and psychiatric disorders among a multiethnic male sample in South Florida. Am. J. Public Health 94, 1603 1609. Gillespie, N., Henders, A., Davenport, T., Hermens, D., Wright, M., Martin, N., Hickie, I., 2013. The Brisbane Longitudinal Twin Study Pathways to Cannabis Use, Abuse and Dependence Project: Current status, preliminary results and future directions. Twin R es. Hum. Genet. Off. J. Int. Soc. Twin Stud. 16, 21 33. Gilreath, T.D., Astor, R.A., Estrada, J.N., Johnson, R.M., Benbenishty, R., Unger, J.B., 2014. Substance use among adolescents in California: A latent class analysis. Subst. Use Misuse. 49, 116 123. Gilson, A.M., Kreis, P.G., 2009. The burden of the nonmedical use of prescription opioid analgesics Pain Med. 10, S89 S100. Goldstein, S., 2008. Report from the National Survey on Drug Use and Health nonmedical stimulant use, other drug use, delinquent b ehaviors, and depression among adolescents J. Atten. Disord. 12, 3 3. Golub, A., Johnson, B.D., 2004. How much do Manhattan arrestees spend on drugs? Drug Alcohol Depend. 76, 235 246. Grant, B.F., Dawson, D.A., 1998. Age of onset of drug use and its ass ociation with DSM IV drug abuse and dependence: results from the National Longitudinal Alcohol Epidemiologic Survey. J. Subst. Abuse 10, 163 173.
144 Griffin, K.W., Botvin, G.J., 2010. Evidence based interventions for preventing substance use disorders in adol escents Child Adolesc. Psychiatr. Clin. N. Am. 19, 505 526. Gruber, E., DiClemente, R.J., Anderson, M.M., Lodico, M., 1996. Early drinking onset and its association with alcohol use and problem behavior in late adolescence Prev. Med. 25, 293 300. Guxen s, M., Nebot, M., Ariza, 2006. Age and sex differences in factors associated with the onset of cannabis use: A cohort study. Drug Alcohol Depend. 88, 234 243. Hall, M.T., H oward, M.O., Mccabe, S.E., 2010 Prescription drug misuse among antisocial youths. J Stud. Alcohol Drugs 71, 917 924. Harden, K.P., Hill, J.E., Turkheimer, E., Emery, R.E., 2008. Gene environment correlation and interaction in peer effects on adolescent alcohol and tobacco use Behav. Genet. 38, 339 347. Harrell, Z.A.T., Broman, C.L., 2009. Racial/ethnic differences in correlates of prescription drug misuse among young adults. Drug Alcohol Depend. 104, 268 271. Harrington, T., Newman, E., 2007. The psychometric utility of two self report measures of PTSD among women substance users. Ad dict. Behav. 32, 2788 2798. Harstad, E., Levy, S., Levy, S., Ammerman, S.D., Gonzalez, P.K., Ryan, S.A., Siqueira, L.M., Smith, V.C., 2014. Attention Deficit/Hyperactivity Disorder and Su bstance Abuse. Pediatrics. 134, 293 301. Hawkins, E.H., 2009. A tal e of two systems : Co occurring mental health and substance abuse disorders treatment for adolescents Annu. Rev. Psychol. 60, 197 227. Hayes, A.F., 2013. Introduction to mediation, moderation, and conditional process analysis : A regression based approach Guilford Press. Heeren, T., Levenson, S., Jamanka, A., Voas, R., 2002. Age of drinking onset, driving after drinking, and involvement in alcohol related motor vehicle crashes. Accid. Anal. Prev. 34, 85 92. Helstrom, A., Bryan, A., Hutchison, K.E., Riggs, P.D., Blechman, E.A., 2004. Tobacco and alcohol use as an explanation for the association between externalizing behavior and illicit drug use among delinquent adolescents. Prev. Sci. Off. J. Soc. Prev. Res. 5, 267 277. Henry, K.L., Knight, K.E., Thornberry, T.P., 2012. School Disengagement as a Predictor of Dropout, Delinquency, and Problem Substance Use during Adolescence and Early Adulthood. J. Youth Adolesc. 41, 156 166.
145 Herman Stahl, M.A., Ashley, O.S., Penne, M.A., Bauman, K.E., Williams, J., Sanchez, R.P., Loomis, K.M., Williams, M.S., Gfroerer, J.C., 2008. Moderation and distress and adolescent substance use : Findings from a national sample J. Adolesc. Health 43, 141 150. Herman Stahl, M.A., Krebs, C.P., Kroutil, L.A., Heller, D.C., 2007. Risk and protective factors for methamphetamine use and nonmedical use of prescription stimulants among young adults aged 18 to 25. Addict. Behav. 32, 1003 1015. Hinshaw, S.P., 1987. On the distinction between attentional deficits/hyperactivity and conduct problems/aggression in child psychopathology. Psychol. Bull. 101, 443 463. Hopfer, C., Salomonsen Sautel, S., Mikulich Gilbertson, S., Min, S. J ., McQueen, M., Crowley, T., Young, S., Corley, R., Sakai, J., Thurstone, C., Hoffenberg, A., Hartman, C., Hewitt, J., 2013. Conduct disorder and initiation of substance use: a prospective longitudinal study J. Am. Acad. Child Adolesc. Psychiatry 52. Ho rton, J., Compton, W., Cottler, L. B. 2000 Reliability of substance use disorder diagnoses among African Americans and Caucasians. Drug Alcohol Depend 57 203 209 Hosmer, D.W., Lemeshow, S., Sturdivant, R.X., 2013. Applied logistic regression John Wile y & Sons. Humphreys, K.L., Eng T, Lee SS, 2013. Stimulant medication and substance use outcomes: A meta analysis. JAMA Psychiatry 70, 740 749. Jacobus, J., McQueeny, T., Bava, S., Schweinsburg, B.C., Frank, L.R., Yang, T.T., Tapert, S.F., 2009. White mat ter integrity in adolescents with histories of marijuana use and binge drinking Neurotoxicol. Teratol. 31, 349 355. Jeffers, A.J., Benotsch, E.G., 2014. Non medical use of prescription stimulants for weight loss, disordered eating, and body image. Eat. B ehav. 15, 414 418. Jeffers, A.J., Koester, S., 2013. Misuse of prescription stimulants for weight loss, psychosocial variables, and eating disordered behaviors. Appetite 65, 8 13. Jessor, R., Jessor, S.L., 1987. Adolescent development and the onset of drinking. A longitudinal study. J Stud Alcohol 36, 27 51. Jessor, R., Jessor, S.L., 1977. The conceptual structure of problem behavior theory, in: problem behavior and psychosocial development: a longitudinal study of youth New York: Academic Press, pp. 19 67.
146 Johansen, M.E., Matic, K., McAlearney, A.S., 2015. Attention Deficit Hyperactivity Disorder medication use among teens and young adults J. Adolesc. Health 57, 192 197. Johnson, E.O., Schultz, L., 2005. Forward telescoping bias in reported age of o nset: an example from cigarette smoking. Int. J. Methods Psychiatr. Res. 14, 119 129. Monitoring the Future national survey results on drug use, 1975 2013: Volume I, Secon dary school students. The University of Michigan, Ann Arbor: Institute for Social Research. Monitoring the Future national survey results on drug use: 1975 2013: Overview, key fin dings on adolescent drug use. The University of Michigan, Ann Arbor: Institute for Social Research. Kalaydjian, A., Swendsen, J., Chiu, W. T., Dierker, L., Degenhardt, L., Glantz, M., Merikangas, K.R., Sampson, N., Kessler, R., 2009. Sociodemographi c predictors of transitions across stages of alcohol use, disorders and remission in the National Comorbidity Survey Replication. Compr. Psychiatry 50, 299 306. Kandel, D.B., 2002. Stages and Pathways of Drug Involvement: Examining the Gateway Hypothesis. Cambridge University Press. Kandel, D., Yamaguchi, K., 1993. From beer to crack: developmental patterns of drug involvement. Am. J. Public Health 83, 851 855. Kaye, S., Darke, S., 2012. The diversion and misuse of pharmaceutical stimulants: what do we kno w and why should we care? Addiction 107, 467 477. Kendler, K.S., Prescott, C.A., Myers, J., Neale, M.C., 2003. The structure of genetic and environmental risk factors for common psychiatric and substance use disorders in men and women. Arch. Gen. Psychia try 60, 929 937. Kent, J.D., Blader, J.C., Koplewicz, H.S., Abikoff, H., Foley, C.A., 1995. Effects of late afternoon methylphenidate administration on behavior and sleep in attention deficit hyperactivity disorder. Pediatrics 96, 320 325. King, K.A., V idourek, R.A., Merianos, A.L., 2013. Sex and grade level differences in lifetime nonmedical prescription drug use among youth J. Prim. Prev. 34, 237 249. King, K.M., Chassin, L., 2007. A prospective study of the effects of age of initiation of alcohol an d drug use on young adult substance dependence J. Stud. Alcohol Drugs 68, 256 265. King, S.M., L acono, W.G., McGue, M., 2004. Childhood externalizing and internalizing psychopathology in the prediction of early substance use. Addiction 99, 1548 1559.
147 Kirby, T., Barry, A.E., 2012. Alcohol as a gateway drug : A study of US 12th graders J. Sch. Health 82, 371 379. Kirisci, L., Tarter, R., Ridenour, T., Reynolds, M., Horner, M., Vanyukov, M., 2015. Externalizing behavior and emotion dysregulation are ind icators of transmissible risk for substance use disorder. Addict B ehaviors 42 57 62. Kollins, S.H., 2008. ADHD, Substance Use Disorders, and psychostimulant treatment current literature and treatment guidelines J. Atten. Disord. 12, 115 125. Lakhan, S.E., Kirchgessner, A., 2012. Prescription stimulants in individuals with and without attention deficit hyperactivity disorder: misuse, cognitive impact, and adverse effects. Brain Behav. 2, 661 677. Lamont, A.E., Woodlief, D., Malone, P.S., 2014. Predict ing high risk versus higher risk substance use during late adolescence from early adolescent risk factors using Latent Class Analysis. Addict. Res. Theory 22, 78 89. Lanza, S.T., Rhoades, B.L., 2013. Latent class analysi s: An alternative perspective on s ubgroup analysis in prevention and treatmen t. Prev. Sci. Off. J. Soc. Prev. Res. 14, 157 168. 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. Lee, T.Y ., 2011. Construction of an integrated positive youth development conceptual framework for the prevention of the use of psychotropic drugs among adolescents. ScientificWorldJournal 11, 2403 2417. Lerner, R.M., Petersen, A.C., Silbereisen, R.K., Brooks Gu nn, J., 2013. The developmental science of adolescence : History through autobiography Psychology Press. the counter drugs in teens: Curr. Opin. Pediatr. 19, 270 274. Levy, S., 2014. Adole scent substance use Pediatr. Ann. 43, 406 407. Liu, J., 2004. Childhood externalizing behavior : Theory and implications J. Child Adolesc. Psychiatr. Nurs. Off. Publ. Assoc. Child Adolesc. Psychiatr. Nurses Inc 17, 93 103. Liu, X., 2003. Cigarette smoki ng, life stress, and behavioral problems in Chinese adolescents. J. Adolesc. Health 33, 189 192. Lord, S., Marsch, L., 2011. Emerging trends and innovations in the identification and management of drug use among adolescents and young adults Adolesc. Med State Art Rev. 22, 649 xiv.
148 Lo, Y., Mendell, N.R., Rubin, D.B., 2001. Testing the number of components in a normal mixture. Biometrika 88, 767 778. Marmorstein, N.R., 2009. Longitudinal associations between alcohol problems and depressive symptoms : Early adolescence through early adulthood Alcohol. Clin. Exp. Res. 33, 49 59. 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 am ong boys J. Abnorm. Child Psychol. 38, 211 224. McCabe, S.E., Boyd, C.J., Teter, C.J., 2009. Subtypes of nonmedical prescription drug misuse. Drug Alcohol Depend. 102, 63 70. McCabe, S.E., Boyd, C.J., Young, A., 2007. Medical and nonmedical use of presc ription drugs among secondary school students J. Adolesc. Health Off. Publ. Soc. Adolesc. Med. 40, 76 83. McCabe, S.E., Brady T. West, Veliz, P., Frank, K.A., Boyd, C.J., 2014. Social contexts of substance use among U.S. high school seniors : A multicohor t national study J. Adolesc. Health 55, 842 844. McCabe, S.E., Cranford, J.A., 2012. Motivational subtypes of nonmedical use of prescription medications: Results from a national study. J. Adolesc. Health Off. Publ. Soc. Adolesc. Med. 51, 445 452. McCab e, S.E., Teter, C.J., Boyd, C.J., 2006. Medical use, illicit use and diversion of prescription stimulant medication J. Psychoactive Drugs 38, 43 56. McCabe, S.E., Teter, C.J., Boyd, C.J., 2004. The use, misuse and diversion of prescription stimulants amo ng middle and high school students. Subst. Use Misuse 39, 1095 1116. McCabe, S.E., West, B.T., 2013. Medical and nonmedical use of prescription stimulants : Results from a national multicohort study J. Am. Acad. Child Adolesc. Psychiatry 52, 1272 1280. McCabe, S.E., West, B.T., Cranford, J.A., Ross Durow, P., Young, A., Teter, C.J., Boyd, C.J., 2011. Medical misuse of controlled medications among adolescents Arch. Pediatr. Adolesc. Med. 165, 729 735. McCabe, S.E., West, B.T., Morales, M., Cranford, J.A ., Boyd, C.J., 2007. Does early onset of non medical use of prescription drugs predict subsequent prescription drug abuse and dependence? Results from a national study. Addict. Abingdon Engl. 102, 1920 1930.
149 McCabe, S.E., West, B.T., Schepis, T.S., Teter, C.J., 2015. Simultaneous co ingestion of prescription stimulants, alcohol and other drugs : A multi cohort national study of U.S. adolescents Hum. Psychopharmacol. 30, 42 51. McCabe, S.E., West, B.T., Teter, C.J., Boyd, C.J., 2012. Co ingestion of prescr iption opioids and other drugs among high school seniors : Results from a national study Drug Alcohol Depend. 126, 65 70. McCauley, J.L., Danielson, C.K., Amstadter, A.B., Ruggiero, K.J., Resnick, H.S., Hanson, R.F., Smith, D.W., Saunders, B.E., Kilpatric k, D.G., 2010. The role of traumatic event history in non medical use of prescription drugs among a nationally representative sample of U.S. adolescents J. Child Psychol. Psychiatry 51, 84 93. McCulloch, Wiggins, R.D., Joshi, H.E., Sachdev, D., 2000. In ternalising and in Britain and the US: Relationships to family resources Child. Soc. 14, 368 383. Merikangas, K.R., He, J., Burstein, M., Swanson, S.A., Avenevoli, S., Cui, L., Benjet, C., Georgiades, K., Swends en, J., 2010a. Lifetime prevalence of mental disorders in US Adolescents: Results from the National Comorbidity Study Adolescent Supplement (NCS A). J. Am. Acad. Child Adolesc. Psychiatry 49, 980 989. Merikangas, K.R., He, J. P., Brody, D., Fisher, P.W., Bourdon, K., Koretz, D.S., 2010b. Prevalence and treatment of mental disorders among US c hildren in the 2001 2004 NHANES. Pediatrics 125, 75 81. Monitoring the Future n ational survey results on drug use, 1975 2014:Volume I, Secondary school students. The University of Michigan, Ann Arbor: Institute for Social Research. Miettunen, J., Murray, G.K., Jones, P.B., Mki, P., Ebeling, H., Taanila, A., Joukamaa, M., Savolainen, J., Trmnen, S., Jrvelin, M. R., Veijola, J., Moilanen, I., 2014. Longitudinal associations between childhood and adulthood externalizing and internalizing psychopathology and adolescent substance use. Psychol. Med. 44, 1727 1738. Molina, B.S.G., Hinsh aw, S.P., Arnold, L.E., Swanson, J.M., Pelham, W.E., Hechtman, L., Hoza, B., Epstein, J.N., Wigal, T., Abikoff, H.B., Greenhill, L.L., Jensen, P.S., Wells, K.C., Vitiello, B., Gibbons, R.D., Howard, A., Houck, P.R., Hur, K., Lu, B., Marcus, S., 2013. Adole scent substance use in the multimodal treatment study of Attention Deficit/Hyperactivity Disorder (ADHD) (MTA) as a function of childhood ADHD, random assignment to childhood treatments, and subsequent medication J. Am. Acad. Child Adolesc. Psychiatry 52 250 263.
150 Molina, B.S.G., Pelham, W.E., 2003. Childhood predictors of adolescent substance use in a longitudinal study of children with ADHD. J. Abnorm. Psychol. 112, 497 507. Molina, B.S.G., Pelham, W.E., Gnagy, E.M., Thompson, A.L., Marshal, M.P., 2007. Attention Deficit/Hyperactivity Disorder risk for heavy drinking and alcohol use disorder is age specific Alcohol. Clin. Exp. Res. 31, 643 654. Morihisa, R. S., Barroso, L. P ., Scivoletto, S. 2007. Labeling disorder: the relationship between conduct problems and drug use in adolescents. Revista Brasileira de Psiquiatria 29 308 314. Morral, A.R., McCaffrey, D.F., Paddock, S.M., 2002. Reassessing the marijuana gateway effect. Addiction 97, 1493 1504. 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 6 2. Mrug, S., Gaines, J., Su, W., Windle, M., 2010. School level substance use : Effects on J. Stud. Alcohol Drugs 71, 488 495. Muthn, L.K., Muthen, B.O., 2007. Mplus Sixth. ed. Muthn & Muthn, Los Angeles, CA. Nakawaki, B., Crano, W.D., 2012. Predicting persistence, and recent onset of nonmedical use of opioids and stimulants Addict. Behav. 37, 716 721. Nargiso, J.E., Ballard, E.L., Skeer, M.R., 2015. A systematic review of risk and protective factors associated with nonmedical use of prescription drugs among youth in the United States: A Social Ecological Perspective. J. Stud. Alcohol Drugs 76, 5 20. Nelson, L.D., Strickland, C., Krueger, R.F., Arbisi, P.A., Patrick, C.J., 2015. Neurobehavioral traits as transdiagnostic predictors of clinical problems Assessment 1073191115570110. Nock, M.K., Kazdin, A.E., Hiripi, E., Kessler, R.C., 2006. Prevalence, subtypes, and correlates of DSM IV Conduct Disorder i n the National Comorbidity Survey Replication. Psychol. Med. 36, 699 710. Nylund, K.L., Asparouhov, T., Muthn, B.O., 2007. Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study Struct. Equ Model. Multidiscip. J. 14, 535 569. Olthuis, J.V., Darredeau, C., Barrett, S.P., 2013. Substance use initiation: The role of simultaneous polysubstance use. Drug Alcohol Rev. 32, 67 71.
151 Pagan, J.L., Rose, R.J., Viken, R.J., Pulkkinen, L., Kaprio, J., D ick, D.M., 2006. Genetic and environmental influences on stages of alcohol use across adolescence and into young adulthood Behav. Genet. 36, 483 497. Palmer, R.S., Corbin, W.R., Cronce, J.M., 2010. Protective strategies: A mediator of risk associated wit h age of drinking onset. Addict. Behav. 35, 486 491. Patton, G.C., Coffey, C., Carlin, J.B., Sawyer, S.M., Lynskey, M., 2005. Reverse gateways? Frequent cannabis use as a predictor of tobacco initiation and nicotine dependence. Addiction 100, 1518 1525. Peleg Oren, N., Saint Jean, G., Cardenas, G.A., Tammara, H., Pierre, C., 2009. Drinking alcohol before age 13 and negative outcomes in late adolescence. alcohol. Clin. Exp. Res. 33, 1966 1972. Piehler, T.F., Vronneau, M. H., Dishion, T.J., 2012. Substan ce use progression from adolescence to early adulthood : Effortful control in the context of friendship influence and early onset use. J. Abnorm. Child Psychol. 40, 1045 1058. Poulin, C., 2007. From attention deficit/hyperactivity disorder to medical stimu lant use to the diversion of prescribed stimulants to non medical stimulant use: connecting the dots. Addict. Abingdon Engl. 102, 740 751. Prescott, C.A., Kendler, K.S., 1999. Age at first drink and risk for alcoholism: a noncausal association. Alcohol. Cl in. Exp. Res. 23, 101 107. Rabiner, D.L., Anastopoulos, A.D., Costello, E.J., Hoyle, R.H., McCabe, S.E., Swartzwelder, H.S., 2009. Motives and perceived consequences of nonmedical adhd medication use by college students are students treating themselves for attention problems ? J. Atten. Disord. 13, 259 270. Rabiner, D.L., Carrig, M., Dodge, K.A., 2013. Attention problems and academic achievement: Do persistent and earlier emerging problems have more adverse long term effects? J. Atten. Disord. Raftery, A.E., 1995. Bayesian model selection in social research. Sociol. Methodol., 25 111 164. Rhoades, H., Winetrobe, H., Rice, E., 2014. Prescription drug misuse among homeless youth Drug Alcohol Depend. 138, 229 233. Riggs, P.D., 2014. Stimulant medication f or ADHD not associated with subsequent substance use disorders. Evid. Based Med. 19, 78 78. Rohrbach, L.A., Milam, L., 2002. Handbook of drug use theory, science, and practice.
152 Roth, A.M., Armenta, R.A., Wagner, K.D., Roesch, S.C., Bluthenthal, R.N., Cue vas Mota, J., Garfein, R.S., 2015. Patterns of drug use, risky behavior, and health status among persons who inject drugs living in San Diego, California: A latent class analysis Subst. Use Misuse 50, 205 214. Sartor, C.E., Agrawal, A., Lynskey, M.T., D uncan, A.E., Grant, J.D., Nelson, E.C., Madden, P.A.F., Heath, A.C., Bucholz, K.K., 2013. Cannabis or alcohol first? Differences by ethnicity and in risk for rapid progression to cannabis related problems in women. Psychol. Med. 43, 813 823. Scalco, M.D., Colder, C.R., Hawk, L.W., Read, J.P., Wieczorek, W.F., Lengua, L.J., 2014. Internalizing and externalizing problem behavior and early adolescent substance use: a test of a latent variable interaction and conditional indirect effects. Psychol. Addict. Beha v. J. Soc. Psychol. Addict. Behav. 28, 828 840. Schachter, D., Tharmalingam, S., Kleinman, I., 2011. Informed consent and stimulant benefits and risks of stimulant medication fo r the treatment of Attention Deficit/Hyperactivity Disorder. J. Child Adolesc. Psychopharmacol. 21, 139 148. Schepis, T.S., Krishnan Sarin, S., 2008. Characterizing adolescent prescription misusers: a population based study. J. Am. Acad. Child Adolesc. Ps ychiatry 47, 745 754. Schinke, S.P., Fang, L., Cole, K.C.A., 2008. Substance use among early adolescent girls: risk and protective factor s. J. Adolesc. Health Off. Publ. Soc. Adolesc. Med. 43, 191 194. Seedall, R.B., Anthony, J.C., 2013. Risk estimates f or starting tobacco, alcohol, and other drug use in the United States: Male female differences and the possibility Drug Alcohol Depend. 133, 751 753. Setlik, J., Bond, G.R., Ho, M., 2009. Adolescent prescri ption ADHD M edication abuse is rising along with prescriptions for these medications. Pediatrics 124, 875 880. Shacham, E., Cottler, L.B., 2010. Sexual behaviors among club drug users: prevalence and reliability. Arch. Sex. Behav. 39, 1331 1341. Simoni Wastila, L., 2000. The use of abusable prescription drugs : The role of gender J. Womens Health Gend. Based Med. 9, 289 297. Simoni Wastila, L., Ritter, G., Strickler, G., 2004. Gender and other factors associated with the nonmedical use of abusabl e prescription drugs. Subst. Use Misuse 39, 1 23. Simoni Wastila, L., Yang, H. W.K., Lawler, J., 2008 Correlates of prescription drug nonmedical use and problem use by adolescents. J. Addict. Med. 2, 31 39.
153 Substance Abuse and Mental Health Services Adm inistration, 2014. Results from the 2013 National Survey on Drug Use and Health: Summary of National Findings (No. HHS Publication No. (SMA)), NSDUH Series H 48. Substance Abuse and Mental Health Services Administration, Rockville, MD. Substance Abuse and Mental Health Services Administration, 2013a. The DAWN Report: Emergency Department Visits Involving Attention Deficit/Hyperactivity Disorder Stimulant Medications. Center for Behavioral Health Statistics and Quality, Rockville, MD. Substance Abuse and Men tal Health Services Administration, 2013b. Drug Abuse Warning Network, 2011: National Estimates of Drug related Emergency Department Visits (No. HHS Publication No. (SMA) 13 4760), DAWN Series D 39. Center for Behavioral Health Statistics and Quality, Rock ville, MD. Substance Abuse and Mental Health Services Administration, 2011. Results 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 He alth Services Administration, Rockville, MD. Substance Abuse and Mental Health Services Administration, 2008. Results from the 2008 National Survey on Drug Use and Health: National Findings (No. HHS Publication No. SMA 09 4434), Office of Applied Studies, NSDUH Series H 36. Substance Abuse and Mental Health Services Administration, Rockville, MD. Sung, H., Richter, L., Vaughan, R., Johnson, P.B., Thom, B., 2005. Nonmedical use of prescription opioids among teenagers in the United States: Trends and correlat es. J. Adolesc. Health 37, 44 51. Sweeney, C.T., Sembower, M.A., Ertischek, M.D., Shiffman, S., Schnoll, S.H., 2013. nonmedical use of prescription adhd stimulants and preexisting patterns of drug abuse J. Addict. Dis. 32, 1 10. Tarter, R.E., Kirisci, L. Mezzich, A., Ridenour, T., Fishbein, D., Horner, M., Reynolds, sequence increase prediction of cannabis use disorder development beyond deviant socialization ? Implications for prevention practice and polic y. Drug Alcohol Depend. 123S1, S72 S78. Taylor, M.J., Merritt, S.M., Brown, C.M., 2012. Perceptions of family caring and its impact on peer associations and drug involvement among rural dwelling African American and White adolesce nts. J. Ethn. Su bst. Abuse. 11, 242 261. Terry drug use among American youth by consumption level, gender, and race/ethnicity : 1976 2005. J. Drug Issues 39, 677 714.
154 Teter, C.J., McCabe, S.E., Cranford, J .A., Boyd, C.J., Guthrie, S.K., 2005. Prevalence and motives for illicit use of prescription stimulants in an undergraduate student sample J. Am. Coll. Health 53, 253 262. Thatcher, D.L., Clark, D.B., 2008. Adolescents at risk for substance use disorder s Alcohol Res. Health 31, 168 176. Tuchman, E., 2010. Women and addiction : The importance of gender issues in substance abuse research J. Addict. Dis. 29, 127 138. Valdez, A., 2014. Rx for inju ry: Adolescent prescription drug misuse J. Emerg. Nurs., Special Disaster issue 40, 497 499. Vande Voort, J.L., He, J. P., Jameson, N.D., Merikangas, K.R., 2014. Impact of the DSM 5 Attention Deficit/Hyperactivity Disorder age of onset criterion in the US adolescent population J. Am. Acad. Child Adolesc. Psyc hiatry 53, 736 744. Van Eck, K., Flory, K., Malone, P.S., 2013. A longitudinal assessment of the associations among response access, attention problems, and aggression during childhood J. Abnorm. Child Psychol. 41, 613 625. Vaughn, M.G., Fu, Q., Perron B., Wu, L. T., 2012. Risk profiles among adolescent nonmedical opioid users i n the United States. Addict. Behav. 37, 974 977. Viana, A.G., Trent, L., Tull, M.T., Heiden, L., Damon, J.D ., Hight, T.L., Young, J., 2012 Non medical use of prescription drug s among Mississippi youth: Constitutional, psychological, and family factors. Addict. Behav. 37, 1382 1388. Visser, S.N., Danielson, M.L., Bitsko, R.H., Holbrook, J.R., Kogan, M.D., Ghandour, R.M., Perou, R., Blumberg, S.J., 2014. Trends in the parent rep ort of health care provider diagnosed and medicated ADHD: United States, 2003 2011. J. Am. Acad. Child Adolesc. Psychiatry 53, 34 46.e2. Visser, S.N., Lesesne, C.A., Perou, R., 2007. National estimates and factors associated with medication treatment for childhood attention deficit/hyperactivity disorder. Pediatrics 119, S99 106. Wagner, F.A., Anthony, J.C., 2002. Into the world of illegal drug us e: Exposure opportunity and other mechanisms linking the use of alcohol, tobacco, marijuana, and cocaine Am. J. Epidemiol. 155, 918 925. 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 (In press) Webb, J.R., Valasek, M.A., North, C.S. 2013. Prevalence of stimulant use in a sample of US medical students. Ann. Clin. Psychiatry Off. J. Am. Acad. Clin. Psychiatr. 25, 27 32.
155 Weiland, B.J., Korycinski, S.T., Soules, M., Zubieta, J. K., Zucker, R.A., Heitzeg, M.M., 2014. Substance abuse risk in emerging adults associated with smaller frontal gray matter volumes and higher externalizing behaviors. Drug Alcohol Depend. 137, 68 75. West, B.T., Teter, C.J., Ross Durow, P., Young, A., Boyd, C.J., 2011. Characteristics associated with the diversion of controlled medications among adolescents. Drug Alcohol Depend. 118, 452 458. Weyandt, L.L., Oster, D.R., Marraccini, M.E., Gudmundsdottir, B.G., Munro, B.A., Zavras, B.M., Kuhar, B., 2014. Pharmacological interventions for adolescents and adults with ADHD: stimulant and nonstimulant medications and misuse of prescription stimulants Psychol. Res. Behav. Manag. 7, 223 249. Wheeler, S.B., 2010. Effects of self esteem and academic performance on adolescent decision making : An examination of early sexual intercourse and illegal substance use J. Adolesc. Health 47, 582 590. Wu, L. T. Pilowsky, D.J., Schlenger, W.E., Galvin, D.M., 2007. Misuse of methamphetamine and prescription stimulants among youths and young adults in the community. Drug Alcohol Depend. 89, 195 205. Wu, L. T., Schlenger, W.E., 2003. Psychostimulant Dependence in a Community Sample. Subst. Use Misuse 38, 221 248. Young, A.M., Glover, N., Havens, J.R., 2012 Nonmedical use of prescription medications among adolescents in the United States: A systematic revi ew. J. Adolesc. Health 51, 6 17. Ystrom, E., Reichborn Kjen nerud, T., Neale, M.C., Kendler, K.S., 2013. Genetic and environmental risk factors for illicit substance use and use disorders : Joint analysis of self and co twin ratings Behav. Genet. 44, 1 13. Zachor, D.A., Roberts, A.W., Hodgens, J.B., Isaacs, J.S., Merrick, J., 2006. Effects of long term psychostimulant medication on growth of children with ADHD. Res. Dev. Disabil. 27, 162 174. Zha, W., 2009. Age of drinking onset, alcohol use disorders, frequent heavy drinking, and unintentionally injuring oneself and others after drinking Pediatrics 123, 1477 1484. Zosel, A., Bartelson, B.B., Bailey, E., Lowenstein, S., Dart, R., 2013. Characterization of adolescent prescription drug abuse and misuse using the Researched Abuse Diversion and Addiction Related Sur veillance (RADARS) System. J. Am. Acad. Child Ad olesc. Psychiatry 52, 196 204
156 BIOGRAPHICAL SKETCH Sonam Ongmu Lasopa received the Bachelor of Arts (Honors) in Psychology from Lady Shri Ram College in 1998, Master of Arts in Applied Psychology from the University of Delhi, South Campus in 2000 and Master of Philosophy in Clinical Psychology from the National Institute of Mental Health and Neuro Sciences (NIMHANS) in 2004. She joined the Department of Psychiatry at Sir Thutob Namgyal Memorial Hospita l under the Department of Healthcare, Human Service and Family Welfare, Government of Sikkim to serve as the first Clinical Psychologist in the state of Sikkim, In dia In 2011, Sonam joined the Epidemiology and Prevention Research Group, Department of Psyc hiatry, College of Medicine, Washington University in St. Louis as a NIH Fogarty Pre d octoral Research Fellow under the mentorship of Dr. Linda. B. Cottler. Subsequently she moved to the Department of Epidemiology, College of Public Health and Health Prof essions and College of Medicine at the University of Florida in 2011 as a Pre doctoral Research Fellow under Dr. Cottler She joined the Ph.D. program in the fall of 2012 with Dr. Cottler serv ing as the chair of the dissertation committee for the work on t he prevalence and risk factors for non medical use of prescription stimulants among youth in the United States. Sonam received the Ph.D. in e pidemiology from the University of Florida in the fall of 2015 and mental health with focus on the assessment of health needs and improv ing access to mental health care service delivery. She has served as both principal and co investigator on government funded grants that assessed the impact of iodine deficiency on c ognitive functioning in young children in two of four districts in Sikkim -an iodine endemic region in the Himalayas Nepal; she was involved in the
157 assessment of health needs in rural Nepal in 2010 and a post earthquake needs assessment following the 7.8 magnitude Nepal earthquake in 2015 as part of services to the Himalayan Family Healthcare Project, a nonprofit organization based in St Louis Missouri. he World Health Organization (WHO) ICD 11 Field Trials for Substance Use and Related Health Conditions at the University of Florida and is part of a w orkgroup on a multisite project to develop a cross cultural measure of life chaos funded by the Fogarty International Centre Recently, Sonam has been involved in a project with the University of Florida Center for Arts in Medicine that pertains to the development of metrics to assess the state level impact of the arts on health in Fl orida. She has published papers and abstracts on assessing health needs in rural Nepal and prescription stimulant misuse and diversion of prescription stimulant medications among youth in the United States. She has receive d awards for her presentations at conferences from the National Institute o n Drug Abuse (NIDA) in the years 2013 to 2015 that include 1) NIDA t ravel award to attend the 103rd Annual meeting of the American Psychopathological Association (APPA) at New York in 2013 ; 2) NIDA Women & Sex/Gend er Junior Investigator t ravel a ward to attend the College of Problems of Drug Dependence (CPDD), Puerto Rico in 2014 and 3) NIDA travel a ward Conference, Arizona in 2015 She also received a scholarship to attend the Summer Institute s of Biostatistics at the University of Washington, Seattle in July, 2015 At the University of Florida, Sonam has been recognized as one of two students among a
158 group of 12 faculty, staff, st udents and alumni who exemplify the Gator Good -a University of Florida mission to change the world for the better through research. S onam has also served as member of a number of professional organizations such as: 1) State Mental Health Authority, Depa rtment of Health Care, Human Service and Family Welfare, Government of Sikkim, India in the years 2006 2008; 2) Life Associate Member of the Indian Psychiatric Society (IPS) since 2004; 3) Student Affiliate Member of the American Psychological Association (APA) in 2013; 4) Student Member of the American Public Health Association (APHA) in 2014. She has been co advisor to two Ph.D. students at the Manipal Institute of Medical Sciences, Sikkim, India and faculty for a Bachelor of Nursing in Psychiatry course at Sir Thutob Namgyal Memorial, Gangtok, Sikkim, India At the University of Florida, Sonam was Teaching Assistant for the course on Principles of Epidemiology under Dr. Cindy Prins in the Department of Epidemiology, College of Public Health and Health Pr ofessions in the years 2012 and 2013. As a member of the District Mental Health Program in Sikkim, India, Sonam has trained several medical officers, nurses, police personnel, teachers and community leaders in the recognition of mental illness to improve a ccess and delivery of mental health services. She has trained counselors appointed under the Sikkim State AIDS Control Society, Government of Sikkim on c ounseling skills for HIV/AIDS and Medical Officers on communication skills and counseling for adolescen ts under the Adolescent Reproductive School Health Program, Government of Sikkim, India.