Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents
(USC Thesis Other)
Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
ENERGY DRINK CONSUMPTION, SUBSTANCE USE AND ATTENTION-DEFICIT/
HYPERACTIVITY DISORDER AMONG ADOLESCENTS
by
Artur Galimov, M.D.
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
PREVENTIVE MEDICINE (HEALTH BEHAVIOR RESEARCH)
December 2021
Copyright 2021 Artur Galimov
ii
Acknowledgements
I would like to give my special regards to my dissertation committee chair Dr. Steven
Yale Sussman, for his dedicated support, guidance, and inspiring me to become a better
researcher. Dr. Sussman continuously provided encouragement and was always willing and
enthusiastic to assist in any way he could through my entire Ph.D. program. I also wish to
express my deepest gratitude to my dissertation committee co-chair Jennifer Unger for offering
constant encouragement and feedback and for her efforts to help me overcome barriers and
challenges during my program. I wish to thank all dissertation committee members whose
assistance was a milestone in the completion of this project: Drs. Lourdes Baezconde-Garbanati,
Jimi Huh, and Reiner Hanewinkel, whose guidance, support, and encouragement have been
invaluable throughout this study. To all IFT-Nord team members for being very supportive and
kind to me during my internship. I also wish to express my gratitude to Ms. Leah Meza for her
feedback, timely advice, and her personal support and empathy throughout the program. Lastly,
nobody has been more important to me in the pursuit of a Ph.D. degree than the members of my
family. I would like to thank my loving and supportive wife Ellie, and my mother, who set me
off on the road to this Ph.D. a long time ago.
iii
Table of Contents
Acknowledgements ......................................................................................................................... ii
List of Tables ................................................................................................................................ vii
List of Figures .............................................................................................................................. viii
Abbreviations ................................................................................................................................. ix
Abstract ........................................................................................................................................... x
Introduction ..................................................................................................................................... 1
Adverse Effects Associated with ED use .................................................................................... 1
Marketing and Sales of EDs ....................................................................................................... 3
Prevalence of ED use Among Adolescents ................................................................................ 4
The Theoretical Rationale for the Studies .................................................................................. 5
Cross-Cultural Similarities and Differences Between U.S. and Germany ................................. 9
Sociodemographic Correlates among Adolescents ................................................................... 12
Perceived Environment and ED use ......................................................................................... 13
Individual Characteristics and ED use ...................................................................................... 14
ED Consumption and Risky Behaviors .................................................................................... 15
ED use and ADHD .................................................................................................................... 16
Limitations to Current ED Research ......................................................................................... 18
Introduction to Dissertation Studies .......................................................................................... 20
Chapter 1: Examining the Association Between Energy Drink Consumption and Tobacco,
Alcohol, and Marijuana Use among German and U.S. Adolescents. ........................................... 24
Introduction ............................................................................................................................... 24
Methods..................................................................................................................................... 26
iv
German Sample Description ................................................................................................. 26
U.S. Sample Description ....................................................................................................... 28
German Sample Measures .................................................................................................... 29
U.S. Sample Measures .......................................................................................................... 32
Data Analysis ........................................................................................................................ 34
Results ....................................................................................................................................... 36
German Study Sample Description ....................................................................................... 36
U.S. Study Sample Description ............................................................................................ 36
Descriptive Analyses ............................................................................................................ 38
Cross-Sectional Analyses ...................................................................................................... 41
Longitudinal Analyses .......................................................................................................... 45
Discussion ................................................................................................................................. 49
Limitations ............................................................................................................................ 53
Conclusions ........................................................................................................................... 54
Chapter 2: Examining the Temporal Associations Between ADHD Symptoms and Energy
Drink Consumption Among German Adolescents. .................................................................. 55
Introduction ............................................................................................................................... 55
Methods..................................................................................................................................... 60
Participants and Procedures .................................................................................................. 60
Measures ............................................................................................................................... 61
Data Analysis ........................................................................................................................ 64
Results ....................................................................................................................................... 65
Study Sample ........................................................................................................................ 65
v
Attrition Analysis .................................................................................................................. 67
Descriptive Analyses ............................................................................................................ 67
Cross-Sectional Analyses ...................................................................................................... 69
Prospective Analyses ............................................................................................................ 69
Supplementary Analyses ....................................................................................................... 72
Discussion ................................................................................................................................. 75
Limitations ............................................................................................................................ 77
Conclusions ........................................................................................................................... 78
Chapter 3: ADHD Symptoms and Substance Use Among German Adolescents: Energy Drink
Consumption as a Risk Factor. ................................................................................................. 80
Introduction ............................................................................................................................... 80
Methods..................................................................................................................................... 83
Participants and Procedures .................................................................................................. 83
Measures ............................................................................................................................... 85
Data Analysis ........................................................................................................................ 88
Results ....................................................................................................................................... 89
Study Sample ........................................................................................................................ 89
Attrition Analysis .................................................................................................................. 90
Descriptive Analyses ............................................................................................................ 92
Multilevel Sem ...................................................................................................................... 92
Discussion ................................................................................................................................. 93
Limitations ............................................................................................................................ 96
Conclusions ........................................................................................................................... 96
vi
Chapter 4: Overall Discussion & Conclusion ............................................................................... 98
Summary of Findings ................................................................................................................ 98
Theoretical Implications ........................................................................................................... 99
Methodological Implications .................................................................................................. 101
Programmatic Implications ..................................................................................................... 102
Overall Conclusion ................................................................................................................. 103
References ................................................................................................................................... 105
vii
List of Tables
Table 1. Comparison of study measures between U.S. and German adolescent samples. ........... 30
Table 2. German and U.S. participant characteristics. .................................................................. 39
Table 3. ED consumption and substance use behavior among German and U.S. adolescents. .... 40
Table 4. German and U.S. participant characteristics separated by lifetime ED use. .................. 42
Table 5. German and U.S. participant characteristics separated by past 30-day ED use. ............ 43
Table 6. Cross-sectional associations between ED consumption and substance use behavior
among German and U.S. adolescents ........................................................................................... 44
Table 7. Predictors of initiating substance use at 12-month follow-up among baseline substance
never users in the German longitudinal sample. ........................................................................... 48
Table 8. Participant baseline characteristics for the total sample and by ED use. ........................ 68
Table 9. Multilevel models examining the cross-sectional associations between study variables
and lifetime/past 30-day ED use. .................................................................................................. 70
Table 10. Multilevel model examining the temporal associations between ED use and ADHD
symptoms. ..................................................................................................................................... 71
Table 11. Multilevel model examining the temporal associations between ED use and ADHD
symptoms among 9–10-years-old children. .................................................................................. 73
Table 12. Multilevel models examining the reversed temporal ordering between ED use and
ADHD symptoms ......................................................................................................................... 74
Table 13. Participant characteristics. ............................................................................................ 90
viii
List of Figures
Figure 1. Heuristic model of the dissertation studies. ..................................................................... 9
Figure 2. Study 1 heuristic model. ................................................................................................ 27
Figure 3. The flow of participants in Study 1. .............................................................................. 37
Figure 4. Association between lifetime ED use at baseline and marijuana use initiation at follow-
up stratified by risk-taking at baseline. ......................................................................................... 46
Figure 5. Study 2 heuristic model. ................................................................................................ 59
Figure 6. The flow of participants in Study 2. .............................................................................. 66
Figure 7. Study 3 heuristic model. ................................................................................................ 84
Figure 8. The flow of participants in Study 3. .............................................................................. 91
Figure 9. Multilevel mediational model. ....................................................................................... 93
ix
Abbreviations
ADHD Attention-Deficit/ Hyperactivity Disorder
AOR Adjusted Odds Ratio
CI Confidence Interval
ED Energy Drink
OR Odds Ratio
SES Socioeconomic Status
TITUS Trends in Tobacco Use Survey
x
Abstract
Energy drinks (EDs), beverages that contain high levels of caffeine in combination with
other ingredients, have become alarmingly popular among children and adolescents around the
globe in recent years. Given this rapid proliferation, it is critical to gain a more systematic
understanding of aspects of ED use behaviors. Two large school-based datasets from two
counties (U.S. and Germany) were used to addresses the following research questions: (1) to
examine the cross-sectional and longitudinal associations between ED consumption and
substance use (cigarettes, e-cigarettes, alcohol, and marijuana) among German and U.S.
adolescents; (2) to examine the temporal associations self-reported ADHD symptoms and ED
use among children and adolescents; (3) to evaluate the mediating role of ED use in the
association between self-reported ADHD symptoms and substance use initiation. Study 1 found
that current ED use was reported by roughly 20% of the participants in both adolescent samples.
Those identified as Hispanic/Latino and multiethnic were more likely to use EDs, while those
identified as Asian were less likely to use them. ED use was cross-sectionally and longitudinally
associated with substance use outcomes (i.e., tobacco, hookah, alcohol, and marijuana). Study 2
demonstrated that ED use is cross-sectionally and prospectively associated with self-reported
ADHD symptoms. Additionally, baseline ED use predicted more frequent ADHD symptoms at
12-months follow-up. Finally, Study 3 showed that ED consumption mediates the association
between self-reported ADHD symptoms and substance use initiation. In sum, the findings of
these studies contribute to a more comprehensive understanding of ED use behavior and its
associated consequences. The research conducted in this dissertation has important implications
for future research methods, theory, and ED prevention and intervention programs, while also
may be beneficial in improving the lives of youth with ADHD symptoms.
1
INTRODUCTION
Energy drinks (EDs) are beverages that contain high levels of caffeine in combination
with other ingredients that are not commonly found in other soft drinks and juices (Harris &
Munsell, 2015; Pomeranz et al., 2013; Vercammen et al., 2019). EDs are marketed to boost
performance, endurance, and mental alertness; however, they should not be confused with sports
or isotonic drinks, which are intended to help athletes to rehydrate and replace electrolytes lost
through exercise (Dillon et al., 2019; Visram et al., 2016). The caffeine content of EDs varies
across brands but typically ranges from 80 mg to 550 mg per 16-oz can. By comparison, most
soft drinks contain approximately 35mg per 12-oz can, while the caffeine content of a 6 oz cup
of coffee is between 77 and 150 mg (Harris et al., 2011; Reissig et al., 2009). In addition to
caffeine, EDs contain other ingredients such as amino acids, sugars or sweeteners, guarana, kola
nut, taurine, ginseng, L-carnitine, herbal supplements, and B vitamins (McCusker et al., 2006;
Seifert et al., 2011). Some of these substances (e.g., guarana, kola nut) may also increase overall
caffeine content, and ginseng is a stimulant as well, yet the US Food and Drug Administration
(FDA) does not require manufacturers to disclose the amounts of these ingredients in nutrition
and supplement facts panels (Markey et al., 2014). As a result, ED users do not know how much
of these substances they are consuming, while independent researchers cannot thoroughly
investigate the interactive effects of ingredients (Harris & Munsell, 2015).
Adverse effects associated with ED use
Given the exponential growth of ED sales coupled with their popularity among the youth,
health professionals have raised concerns regarding the potential detrimental health effects of
EDs for children, adolescents, and young adults. Growing evidence suggests that stimulants
2
(such as caffeine, guarana, ginseng, and kola nut) and other ingredients contained in EDs can
elevate the blood pressure, heart rate and interfere with internal calcium reabsorption, which may
lead to various physical adverse effects (Sanchis-Gomar et al., 2015; Seifert et al., 2011; Somers
& Svatikova, 2020). A few studies have demonstrated that ED use is associated with
cardiovascular problems. For instance, ED use among adolescents and children has been linked
with hypertension (Seifert et al., 2011; Usman & Jawaid, 2012). The results of a randomized
clinical trial showed that Red Bull© consumption (355 mL) among twenty-five healthy young
adults (20-31 years) was associated with elevated blood pressure and a lower cerebral blood flow
velocity (Grasser et al., 2014). Two studies reported an ST-elevation (on electrocardiogram) and
coronary vasospasms in 17- and 19-year-old adolescents after overconsumption of EDs (Scott et
al., 2011; Wilson et al., 2012). Moreover, adolescents who use EDs have been diagnosed with
arrhythmias and palpitations (Di Rocco et al., 2011; Gunja & Brown, 2012; Sanchis-Gomar et
al., 2015). A review study by Ehlers and colleagues reported that a high intake of EDs (about 30
oz) was associated with moderate to severe cardiovascular problems among young adults, such
as prolonged QT-interval (on electrocardiogram) and palpitations (Ehlers et al., 2019).
Given the high caffeine concentration in EDs, a growing body of literature suggests that
adolescent ED use is associated with caffeine dependence and addiction (American Academy of
Pediatrics, 2011; Pomeranz et al., 2013). Caffeine binds to cell membranes in place of adenosine
(inhibitory neurotransmitter), which results in dysregulation of normal physiological processes
(Pomeranz et al., 2013). Caffeine abstinence can result in the following withdrawal symptoms:
headaches, anxiety, apathy, weakness, tremor, nausea, and vomiting (Fredholm et al., 1999).
Several studies have demonstrated that ED consumption is associated with reduced sleep quality,
decreased daytime functioning, fatigue, headaches, insomnia, and difficulty breathing among
3
adolescents (Bashir et al., 2016; Hammond et al., 2018; Koivusilta et al., 2016; Park et al., 2016)
and young adults (Faris et al., 2017; Larson et al., 2015; Richards & Smith, 2016b; Silverman et
al., 1992; G. Trapp et al., 2020). Other adverse health effects associated with adolescent ED
consumption include gastrointestinal (i.e., gastric disturbance and nausea) and neurological
symptoms (i.e., jitteriness, seizures, and tremors) (Gunja & Brown, 2012; Hammond et al., 2018;
Pennington et al., 2010).
Marketing and Sales of EDs
Although EDs originated in Europe and Asia in the 1960s, they became widely available
to the public with the emergence of Red Bull© in Austria in 1987 and the U.S. in 1997 (Reissig
et al., 2009). Since then, the ED industry has been rapidly growing and becoming a multibillion-
dollar market. In 2015 total worldwide sales of EDs reached $50 billion and are expected to top
$60 billion by 2021 (AIM Market Insight, 2015). Initially, ED companies targeted athletes
offering improvement in energy and stamina, yet more recently, the market (in the U.S. and
globally) became more youth-oriented (Hammond & Reid, 2018; Harris, 2013; Harris &
Munsell, 2015; Stacey et al., 2017). Marketing and advertising strategies are rapidly evolving
and reflect complex youth lifestyles. Television has long been the staple of advertising to
children and youth (Calvert, 2008); one study reported that children view approximately 40,000
advertisements each year (Kunkel & Castonguay, 2012). One study has shown that the average
number of ED television advertisements seen by adolescents aged 12-17 years doubled since
2008 and reached 165 ads (on average one ad every two days) in 2012 (Harris, 2013). Moreover,
these advertisements were placed on networks that youth under 18 years watched relatively more
often (i.e., youth-oriented TV channels) than adults (Hammond & Reid, 2018; Harris, 2013;
4
Stacey et al., 2017). Other youth-oriented marketing strategies include internet advertisements,
product placement in video games and social media, as well as sponsorship of concerts and
sporting events (where they provide product samples) that appeal to adolescents (e.g.,
snowboarding, motocross) (Emond et al., 2015; Hammond & Reid, 2018; Harris & Munsell,
2015; Harris et al., 2011).
Prevalence of ED use among adolescents
The prevalence rates of ED use among children and adolescents are identical across
different studies. For instance, Miller and colleagues reported that 64.3% of U.S. adolescents
aged 13-17 years (n=1,032) consumed EDs in their lifetime, while 21.8% were past 30-day users
(Miller et al., 2018). Azagba et al. found that 62.2% of Canadian adolescents (n= 8210, mean age
15.2 years) were past 12-month ED users, while 20.0% reported using them in the past 30-days
(Azagba et al., 2014). A study conducted in Australia found that 51.2% of adolescents (n= 3688,
mean age 13.6 years) used EDs in their lifetime, while 23.9% were past 30-day users (Trapp et
al., 2020). According to the European Food Safety Authority, the lifetime prevalence of ED use
among European adolescents aged 15-18 years is 68%, and it is 18% among 6-10 years old
children, with the lowest rates being reported in Greece (48%) and the highest in Belgium (85%)
(Zucconi et al., 2013). Finally, Galimov et al. (2019) reported 61.7% lifetime ED use among
German adolescents aged 9-18 years and 21.4% past 30-day use. The average age of initial ED
consumption reported in the recent studies ranges between 10-15 years (Costa et al., 2016; Reid
et al., 2015; Trapp et al., 2020).
5
The theoretical rationale for the studies
It was demonstrated that ED use behavior might share similar characteristics and qualities
with other substance or behavioral addictions, which are often described by appetitive effects
(appetitive need and satiation), preoccupation, loss of control, and negative consequences
(Sussman, 2017). Nonetheless, theoretical background literature on ED use behavior is limited,
nor is there a unique theory that that has been utilized to explain ED use among youth. This
dissertation integrates multiple theories to explain the associations between ED consumption,
substance use, and attention deficit hyperactivity disorder (ADHD).
A few studies (Galimov et al., 2019; Miller, 2008a) have utilized Problem Behavior
Theory in an attempt to explain the patterns of ED use behavior (Jessor, 1991). This theory
utilizes three interactive systems of psychosocial influence (i.e., personality, perceived
environment, and behavior) to explain substance use and other unconventional lifestyles. Within
each of these systems, the exploratory variables act as instigators of or controls against problem
behavior. For instance, factors like family/peers’ control and support, the value on academic
achievement, and attitudinal intolerance of deviance are protective from problem behaviors,
while family/peer models of problem behavior, vulnerability risk factors (i.e., stress, depression,
low self-esteem) and opportunity risk factors (i.e., availability of drugs at home/neighborhood)
are instigating problem behaviors (Jessor et al., 2003). Substance use and other risky behaviors
are often learned together (as a syndrome of problem behavior) and usually co-exist in the social
ecology of adolescent life; as a result, involvement in one problem behavior increases the
likelihood of involvement in other problem behavior (Donovan et al., 1991; Jessor, 1991). In line
with this theory, if ED consumption is associated with subsequent substance use initiation, ED
6
use may be merely a marker for risk-taking adolescents, reflecting a common liability for
substance use in general (see Figure 1).
It is also possible that ED consumption can act as a “gateway” to the use of other
substances. The “gateway hypothesis” concept was first introduced by Denise Kandel and
colleagues in 1975, who evaluated the systematic trends in substance use sequences among youth
(Kandel, 1975). This theory suggests that there are developmental stages of drug use among
youth. A few studies have suggested that there is a clearly defined sequence of drug use onset
starting with licit substances in adolescence (cigarettes, alcohol) and progression to illicit drugs
use (from marijuana to cocaine, methamphetamine, and heroin) in adulthood (Kandel, 1975;
Kandel et al., 1992; Wagner & Anthony, 2002). Numerous studies have tested this hypothesis by
controlling for several confounders; nonetheless, the underlying mechanisms of the observed
associations remain controversial (Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli,
2016; Seidel et al., 2021; Vanyukov et al., 2012). For instance, a few studies suggested that e-
cigarettes can develop into a gateway to nicotine addiction and combustible tobacco use
(Leventhal et al., 2015; Martinelli et al., 2021), while other studies showed that there is a reverse
relation between tobacco smoking and initiation of e-cigarette use and argued that the gateway
hypothesis had not been proven (East et al., 2018; Kozlowski & Warner, 2017). Nonetheless,
given the stimulating effect of ingredients contained in EDs and their widespread availability at
different retailers, EDs may well be another licit substance that acts as a “gateway” to licit
substances (i.e., cigarettes, e-cigarettes, hookah, alcohol) and illicit drugs.
Further, neurobiological/dysregulation theories also may support the “ED gateway”
perspective. A growing body of the literature suggests that almost all drugs of abuse, as well as
other addictive behaviors (gambling, overeating), increase dopamine activity, release, and
7
utilization in the nucleus accumbens (Brody et al., 2009; Krause et al., 2002; Sussman, 2017;
Sussman & Ames, 2008). The nucleus accumbens is a structure of the mesolimbic dopamine
pathway that involves a projection of dopamine neurons from the midbrain’s ventral tegmental
area to the nucleus accumbens and then on to the prefrontal as well as the anterior cingulate
cortex. It was demonstrated that the dopamine neurons located in the nucleus accumbens encode
and process the proximal stimuli that are associated with the rewarding experiences, including
drug or other addictive behavior stimuli (Sussman, 2017; Sussman & Ames, 2008). Further, it
was shown that repeated sensitization to one specific substance (that stimulates the release of
dopamine) enhances the response to that substance (i.e., it makes the substance more rewarding)
as well as to other drugs that act at the same neurobiological sites, including ones that may have
never been previously consumed (Hellberg et al., 2019; Robinson et al., 1985; Sussman, 2017;
Temple, 2009). Consequently, consumption of one drug can act as a “gateway” to the use of
other drugs. The caffeine contained in ED is well known for its stimulating effect on mesolimbic
reward pathways of the brain (Nall et al., 2016; Nehlig et al., 1992; Solinas et al., 2002).
Therefore, in line with the cross-sensitization hypothesis (Hellberg et al., 2019; Robinson et al.,
1985; Sussman, 2017; Temple, 2009), repeated ED use could cause dysregulation in mesolimbic
structures of the brain. As a result, ED “sensitized” individuals may be more prone to experience
pathological motivation for other substances (see Figure 1).
Additionally, a few studies suggested that dopamine reinforcement dysfunction and
abnormalities in the structure of critical brain regions related to dopamine are primary causes of
ADHD (Swanson et al., 2007; Volkow et al., 2009). Hence, a few researchers proposed that
adolescents with ADHD may choose to self-medicate with nicotine, alcohol, and other
stimulating substances to manage their symptoms. This pattern is an example of the self-
8
medication hypothesis (Krause et al., 2002; Lawrence et al., 2002; Levin et al., 2006). EDs can
also cause a strong stimulating effect, attributed to the interaction between the various
ingredients contained in these beverages, such as caffeine, guarana, ginseng, and kola nut
(Reissig et al., 2009; Vercammen et al., 2019). Moreover, unlike alcohol or tobacco, EDs do not
have a minimum sale age and can be purchased at any convenience or grocery store. Given the
increasing popularity of EDs among minors, ADHD adolescents may use them to manage their
symptoms. Thus, it is possible that having ADHD symptoms among adolescent ED non-users is
associated with future risk of ED use initiation. Furthermore, evidence suggests that alterations
in dopamine function transfer may give rise to many symptoms of ADHD (Tripp & Wickens,
2009). Thus, the dysregulation of mesolimbic function caused by ED use in children and
adolescents with ADHD symptoms may further exacerbate the signs and symptoms of ADHD.
The heuristic model of this dissertation proposal, which integrates the problem behavior
(Donovan et al., 1991; Jessor, 1991; Jessor et al., 2003), gateway hypothesis (Kandel, 1975;
Kandel et al., 1992; Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli, 2016; Vanyukov et
al., 2012; Wagner & Anthony, 2002), cross-sensitization hypothesis (Hellberg et al., 2019;
Robinson et al., 1985; Sussman, 2017; Temple, 2009), and self-medication hypothesis (Krause et
al., 2002; Lawrence et al., 2002; Levin et al., 2006) is depicted on Figure 1.
9
Figure 1. Heuristic model of the dissertation studies.
Notes: green paths correspond to Study 1 hypotheses; blue paths correspond to Study 2 hypotheses; red paths
correspond to Study 3 hypotheses.
Cross-cultural similarities and differences between U.S. and Germany
Previous studies have emphasized the importance of the cultural and social environment
in understanding and explaining the prevalence and patterns of adolescent substance use
(Bauman & Phongsavan, 1999; Link, 2008). Given the global proliferation of EDs in recent
years, it is also critical to gain a more systematic understanding of aspects of ED use behavior.
The respective social and cultural environments can directly or indirectly shape substance and
ED use behaviors by altering factors correlated with these behaviors. Hence, this dissertation
evaluated the associations between ED consumption, substance use, and ADHD by analyzing the
10
data from two large school-based adolescent samples: U.S. and German. Adolescents living in
the U.S. and Germany have many similarities regarding their daily lives. Moreover, both
countries are well-developed and share similar macroeconomic characteristics, yet they represent
distinct cultures in terms of commonly examined frameworks, while differences in laws and
regulations also exist. Although it may be informative to compare vastly different countries
(Eastern vs. Western), it is critical to evaluate the differences between cultures that bear
significant similarities (i.e., U.S. and Germany), such as access to health care, political (e.g.,
democratic values) and economic systems.
U.S. and Germany differ with regards to the minimum age of sale of tobacco products
and minimum legal drinking age. For instance, in the U.S., it is illegal for a retailer to sell any
tobacco products - including cigarettes, cigars, and e-cigarettes - to anyone under 21 years (US
Food and Drug Administration, 2020); on the other hand, the minimum age of sale of tobacco
products in Germany is 18 years (Nuyts et al., 2020). Similarly, the minimum legal drinking age
in the U.S. is 21 years (Centers for Disease Control and Prevention, 2020), while the minimum
legal drinking in Germany is 16 years for undistilled alcoholic beverages (such as wine and beer)
and 18 years for spirits (European Union Agency for Fundamental Rights, 2018). Not
surprisingly, the prevalence of smoking and alcohol use behaviors differs dramatically across the
two countries. For instance, the Monitoring the Future U.S. national survey results reported
18.2% past 30-day alcohol use prevalence and 3.7% cigarette use prevalence among 8th, 10th,
and 12th graders in 2019 (Johnston et al., 2020). On the other hand, 6.0% past 30-day cigarette
smoking and 30.0% past 30-day alcohol use were reported among German adolescents (mean
age 13-14 years; de Bruijn et al., 2016; Hansen et al., 2018).
11
In this context, it is crucial to note that the cultural message sent by the lower age limit
may increase adolescents’ perceived safety of substance behaviors and impact their beliefs
regarding the risk of negative (legal and personal) consequences, which in turn could increase
their use of substances (Link, 2008; Weinstein, 1982). As follows, growing evidence suggests
that higher minimum legal age for alcohol and tobacco use significantly helps to reduce youths’
ability to acquire cigarettes from slightly older peers and contributes to a more significant decline
in youth alcohol consumption and smoking (Bonnie et al., 2015; Nuyts et al., 2020; Schneider et
al., 2016; Wagenaar & Toomey, 2002). Thus, substance use prevalence in the German adolescent
sample may be higher than in the U.S. sample. Neither Germany nor the U.S. has adopted
policies restricting sales of ED; hence ED use prevalence may be roughly identical across two
adolescent samples.
Similarities and differences in value systems between these two countries are also
reflected in Hofstede’s work, where he describes five universal dimensions of culture:
individualism, masculinity, power-distance, uncertainty avoidance, and long-term orientation
(Hofstede, 1994). Individualism, a cultural value in many industrialized countries (i.e., U.S.), is a
tendency to give priority to personal interests, individual freedom, emotional detachment, self-
sufficiency, and fulfillment of personal needs (Triandis et al., 1990). In contrast, German culture
is considered more collectivistic than the U.S. (Hofstede, 1994) and values interdependence,
emotional closeness, group achievement, and cooperation (Triandis et al., 1990). In this line,
German adolescents high in collectivism may seek to adjust to society's norms and expectations,
while U.S. adolescents high in individualism may seek to rebel against them (Unger et al., 2002).
Thus, it is possible that peer influence would be a more powerful predictor of substance and ED
12
behavior in the German sample, while rebelliousness and risk-taking (Grichnik, 2008; Lafreniere
et al., 2013) would be more strongly related to substance and ED use in the U.S. sample.
Finally, while the U.S. is increasingly becoming a multicultural and ethnically diverse
society (Unger et al., 2002; Unger et al., 2001), the ethnic composition of the German population
is less diverse (Schenk et al., 2007). For instance, it was shown that children and adolescents
with migration backgrounds account for only 17.1% of all children and adolescents in Germany
(Schenk et al., 2007). Moreover, most of those with migration backgrounds are migrants from
other European countries (Froehlich et al., 2019; Schenk et al., 2007). On the other hand, in
ethnically diverse areas of the U.S., such as California, children, and adolescents of many
different ethnic backgrounds interact with one another daily and influence one another’s attitudes
and beliefs (Unger et al., 2002; Unger et al., 2001). A few studies have noted ethnic differences
with regards to substance use behaviors, and evidence suggests that the prevalence of cigarette,
e-cigarette, and alcohol use is higher among White and Hispanic/Latino adolescents than among
Black and Asian adolescents (Barrington-Trimis et al., 2019; Chen & Unger, 1999; Kann et al.,
2018). Further, White and Hispanic/Latino youth are more likely to initiate cigarettes through e-
cigarettes, while Black adolescents are more likely to initiate cigarettes through cigars (Stokes et
al., 2020). Thus, the association between ED consumption and substance use in the U.S. sample
may vary by ethnicity, while the results in the German sample are more homogeneous.
Sociodemographic correlates among adolescents
Sociodemographic correlates associated with ED use are highlighted in the literature.
Most studies report that boys are more likely to consume EDs than girls (Azagba et al., 2014;
Kaur et al.; Visram et al., 2016). Moreover, one longitudinal study among adolescents aged 9-19
13
years demonstrated that boys have higher odds of initiating ED use over 12-months compared to
girls (Galimov et al., 2019). Growing evidence suggests that older adolescents are more likely to
use EDs than younger ones (Galimov et al., 2019; Gallimberti et al., 2013; Leal & Jackson,
2018; Sampasa-Kanyinga et al., 2020). Additionally, a few studies have demonstrated that those
adolescents from a single-parent family with lower levels of parental monitoring and less
educated parents are more likely to use EDs (Miyake & Marmorstein, 2015; Terry-McElrath et
al., 2014).
The patterns of ED use with regards to the ethnic background are less consistent. One
U.S. study has shown that African American college students were less likely to report ED use
compared to their White peers (Miller, 2008b), while a few other studies conducted among U.S.
adolescents demonstrated that the converse was true (Miller et al., 2018; Park et al., 2012).
Further nationwide studies involving large youth samples are needed to better understand the
variation in ED consumption between different ethnic groups.
Perceived environment and ED use
Perceived environment characteristics associated with ED use include ED advertisement
exposure, peer ED use, and parental controls. A few studies have shown that ED manufacturers
have increased youth-oriented advertising on the Internet and primarily promote their products
on TV channels that likely appeal to adolescents (Emond et al., 2015; Harris, 2013). Not
surprisingly, one study found that more frequent ED advertisement exposure is associated with
more frequent ED use and is predictive of ED use initiation (Galimov et al., 2019). Peer ED use
has also been linked with higher ED use frequency, yet was not predictive of ED use initiation
14
(Galimov et al., 2019). Finally, evidence suggests that low levels of parental monitoring are
associated with greater ED consumption among adolescents (Miyake & Marmorstein, 2015).
Individual Characteristics and ED use
There is accumulating evidence that ED use is associated with sensation-seeking and
risk-taking (Azagba et al., 2014; Emond et al., 2014; Galimov et al., 2019; Hamilton et al.,
2013). Additionally, two longitudinal studies demonstrated that higher levels of sensation-
seeking are predictive of ED use initiation (Galimov et al., 2019; Miyake & Marmorstein, 2015).
Moreover, several cross-sectional studies conducted among adolescents have linked ED
consumption to mood disorders, such as anxiety, depression, and suicidal ideation, plan or
attempt (Al-Shaar et al., 2017; Masengo et al., 2020; Park et al., 2016; Richards & Smith,
2016b).
Growing evidence suggests that ED consumption is associated with relatively poor
academic performance (Galimov et al., 2019) and unhealthy lifestyle behaviors (Faris et al.,
2017). More specifically, several studies demonstrated that regular ED use (i.e., past 30-day, past
week) is associated with poor dietary habits, such as regular junk food or fast-food consumption
(Galimov et al., 2019; Poulos & Pasch, 2015; Richards & Smith, 2016a). Not surprisingly,
several studies found that regular ED use is associated with elevated body mass index (Galimov
et al., 2019; Poulos & Pasch, 2015; Reid et al., 2015).
A few studies have identified taste as the primary factor motivating youth to use and
purchase EDs (Bunting et al., 2013; Costa et al., 2014; Visram et al., 2016; Visram et al., 2017).
Additionally, European studies reported that many adolescents view EDs as a cheaper alternative
to soft drinks (Costa et al., 2014; Visram et al., 2016; Visram et al., 2017). For instance, two
15
studies have reported that some ED brands are priced as low as 0.5 EUR (0.5 USD) or 0.35£
(0.5USD) per can (Nowak & Jasionowski, 2015; Visram et al., 2017). The most common reasons
encouraging adolescent ED use include compensation for insufficient sleep, the necessity to
increase physical performance and boost energy, and the desire to improve academic
performance and decrease fatigue (Costa et al., 2014; Malinauskas et al., 2007; Visram et al.,
2016). Finally, Galimov et al. (2019) reported that curiosity to try ED is one of the key factors
associated with ED use initiation.
ED Consumption and Risky Behaviors
Risky behaviors associated with ED use are well documented in the literature. For
instance, consistent with the Problem Behavior Theory (Donovan et al., 1991; Jessor, 1991;
Jessor et al., 2003), a few cross-sectional studies demonstrated that ED consumption is correlated
with tobacco, alcohol, marijuana, and other illicit drug use, as well as with sexual risk-taking and
violence (Azagba et al., 2014; Emond et al., 2014; Gallimberti et al., 2013; Hamilton et al., 2013;
Miller, 2008a; Terry-McElrath et al., 2014). Additionally, two longitudinal studies have shown
that ED use frequency at baseline is associated with higher alcohol use frequency at 12-month
(Choi et al., 2016) and 16-month (Miyake & Marmorstein, 2015) follow-up. Nonetheless, these
two studies were conducted among current adolescent substance users and failed to control for a
current cigarette, e-cigarette, hookah, or marijuana use, and other important confounders such as
peer substance use, vulnerability risk factors (i.e., stress, depression), and SES. Thus, whether
ED is a risk factor for initiating cigarettes, e-cigarettes, hookah, alcohol, marijuana, and other
substances remain unclear.
16
ED Use and ADHD
ADHD is a neurodevelopmental disorder characterized by persistent symptoms of
hyperactivity, impulsivity, and difficulty sustaining attention (American Psychiatric Association,
2013; Hughes et al., 2015). ADHD is one of the most common psychiatric conditions in children
and adolescents. A recent systematic review that included 179 ADHD prevalence estimates
reported an overall pooled prevalence of 7.2% (95% CI: 6.7 -7.8) among youth (Thomas et al.,
2015). Moreover, there is evidence suggesting that this rate is possibly increasing in certain
populations (Nyarko et al., 2017; Visser et al., 2010). ADHD tends to run in families, and its
etiology involves a combination of neurodevelopmental, environmental, and genetic components
(Thapar et al., 2013).
ADHD is a major clinical and public health problem because of its associated morbidity
and disability in children, adolescents, and adults (Goldman et al., 1998). ADHD youth often
suffer from school problems. For instance, growing evidence suggests that adolescents with a
formal ADHD diagnosis, as well as those with self-reported ADHD symptoms, experience
educational problems, report poor academic performance and peer relationship difficulties (Barry
et al., 2002; Diamantopoulou et al., 2005; Gardner & Gerdes, 2015; Rapport et al., 1999; Strine
et al., 2006). Moreover, a wide range of comorbid psychiatric disorders are associated with
ADHD (in clinical and community samples), including oppositional defiant disorder, conduct
disorder, anxiety, and depression (Biederman, 2005; Jensen et al., 2001; Pan & Yeh, 2017; Yen
et al., 2007). Serval longitudinal studies have demonstrated that adolescents with ADHD are
more likely to experience an early onset of substance use disorders than non-ADHD individuals
(Biederman et al., 1997; Katusic et al., 2005; Wilens et al., 2011). Moreover, a comparative
meta-analysis of thirteen studies has shown that childhood ADHD is associated with drug and
17
alcohol disorders in adulthood and with nicotine use in adolescence (Charach et al., 2011). While
most of these studies have been conducted among clinical adolescent samples, growing evidence
suggests that self-reported ADHD symptoms are associated with substance use behavior in
general population samples. For instance, it was shown that self-reported ADHD symptoms
among adolescents with no formal ADHD diagnosis are associated with substance (alcohol,
nicotine, marijuana, cocaine) use behaviors (De Alwis et al., 2014; Kessler et al., 2012).
Given that dopamine reinforcement dysfunction is one of the primary causes of ADHD
(Swanson et al., 2007), an association between ADHD and substance use is not surprising. There
is accumulating evidence suggesting that almost all drugs of abuse increase dopamine activity
(Brody et al., 2009; Krause et al., 2002; Sussman, 2017; Sussman & Ames, 2008), hence,
consistent with the self-medication hypothesis (Krause et al., 2002; Lawrence et al., 2002; Levin
et al., 2006), adolescents with ADHD may choose to self-medicate with nicotine, alcohol, and
other stimulating substances to manage their symptoms. In line with this hypothesis, it is also
plausible that ADHD adolescents may use other stimulating licit substances, such as EDs, to
cope with their symptoms. In such a way, one cross-sectional study discovered an association
between ED consumption and self-reported hyperactivity/inattention symptoms (Schwartz et al.,
2015). Additionally, another study suggested that children with teacher-reported ADHD
symptoms may be at higher risk of ED use than to those without ADHD-like symptoms
(Marmorstein, 2016). Nonetheless, the temporal associations between ED use and ADHD
symptoms among adolescents are not yet fully understood.
18
Limitations to Current ED Research
Despite growing concern for ED consumption among adolescents, several gaps in the
literature remain. First, most of the research investigating the relationship between ED
consumption and substance use (i.e., tobacco, e-cigarette, alcohol, and marijuana use) has been
limited to cross-sectional studies. While two longitudinal studies have documented that the initial
frequency of ED consumption predicted an increase in the frequency of alcohol use in the future
(Choi et al., 2016; Miyake & Marmorstein, 2015), these two studies were conducted among
current adolescent substance users and failed to control for a current cigarette, e-cigarette,
hookah or marijuana use, as well as other important confounders such as peer substance use,
vulnerability risk factors (i.e., stress, depression), and SES. Thus, whether the use of EDs among
substance non-users is associated with future risk of initiating tobacco, e-cigarette, alcohol, and
marijuana use remain unknown.
Second, while it is plausible that ED consumption is a risk factor for initiating the use of
other substances, little is known about the mechanisms that would account for this potential link.
In line with the cross-sensitization (Hellberg et al., 2019; Robinson et al., 1985; Sussman, 2017;
Temple, 2009) and gateway hypotheses (Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli,
2016; Seidel et al., 2021; Vanyukov et al., 2012) ED use may serve as a gateway drug to other
substances. On the other hand, consistent with the Problem Behavior Theory (Donovan et al.,
1991; Jessor, 1991; Jessor et al., 2003), it could be that using EDs is simply a marker for risk-
taking individuals in general. Nonetheless, additional research is needed to better understand
these mechanisms.
Third, questions remain regarding the temporal associations between ADHD symptoms
and ED use, as current research is limited to one cross-sectional and one longitudinal study
19
conducted among children (Marmorstein, 2016; Schwartz et al., 2015). Although the longitudinal
study demonstrated bi-directional associations between ADHD-like symptoms and ED intake,
these findings were limited by the study's small sample size (n=134), limited covariates list, and
children population (mean age = 11.9). Previous studies have demonstrated that the average age
of initial ED use ranges between 10-15 years (Costa et al., 2016; Reid et al., 2015; Trapp et al.,
2020). Thus, it is critical to include adolescent participants when examining the temporal
relationship between ADHD and ED use. Moreover, this study failed to control for current
substance use (i.e., cigarette, e-cigarette, hookah, alcohol, and marijuana), as well as other
important confounders such as peer ED use, academic achievement, and vulnerability risk factors
(i.e., stress, depression), which likely have affected the results. This highlights the need for
further prospective studies that would incorporate a representative community sample, older age
groups, and control for important covariates associated with ADHD and ED use. Dopamine
reinforcement dysfunction is one of the primary causes of ADHD (Swanson et al., 2007); hence,
consistent with the self-medication hypothesis (Krause et al., 2002; Lawrence et al., 2002; Levin
et al., 2006), adolescents with ADHD may choose to self-medicate with EDs to cope with their
symptoms. In this line, it is plausible that having ADHD symptoms among adolescent ED non-
users is associated with future risk of ED use initiation. Furthermore, evidence suggests that
alterations in dopamine function transfer may give rise to many symptoms of ADHD (Tripp &
Wickens, 2009); thus, dysregulation of mesolimbic function caused by ED use in children and
adolescents with ADHD symptoms may further lead to exacerbation of signs and symptoms of
ADHD. Additionally, given the prospective association between ADHD symptoms (clinical and
subclinical) and substance use among adolescents (Biederman et al., 1997; De Alwis et al., 2014;
Katusic et al., 2005; Kessler et al., 2012; Wilens et al., 2011) and if ED use is associated with
20
substance use initiation, it is also plausible that ED consumption mediates the association
between ADHD symptoms and tobacco, e-cigarette, alcohol, and marijuana use initiation (Baron
& Kenny, 1986; Fairchild & MacKinnon, 2009).
Introduction to Dissertation Studies
This dissertation extends the previous work and fills identified gaps in the literature by
evaluating the patterns of ED use behavior and examining the associations between ED
consumption, substance use, and ADHD in two adolescent samples (U.S. and German). The
proposed dissertation is grounded on (see Figure 1) the Problem Behavior Theory (Donovan et
al., 1991; Jessor, 1991; Jessor et al., 2003) gateway hypothesis (Kandel, 1975; Kandel et al.,
1992; Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli, 2016; Vanyukov et al., 2012;
Wagner & Anthony, 2002), self-medication theory (Krause et al., 2002; Lawrence et al., 2002;
Levin et al., 2006), and cross-sensitization hypothesis (Hellberg et al., 2019; Robinson et al.,
1985; Sussman, 2017; Temple, 2009).
Study 1 of the proposed dissertation estimates the prevalence of ED use and describes the
epidemiology by demographic characteristics in U.S. and German adolescent samples. Further,
cross-sectional associations between ED consumption and substance use (tobacco, e-cigarette,
alcohol, and marijuana) are tested. Next, in the German sample of baseline substance non-users,
we evaluate whether baseline ED consumption is associated with the risk of initiating tobacco, e-
cigarette, alcohol, and marijuana use in 12 months. Additionally, whether risk-taking moderates
these associations is examined. Study 2 examines a longitudinal association between ADHD
symptoms frequency and ED consumption. Since the temporal relationship between these two
factors is unknown, we test whether the baseline frequency of ADHD symptoms among ED non-
21
users at baseline predicts the ED use initiation in 12 months. The reversed temporal ordering is
also tested to examine the evidence for potential reverse causation. Finally, Study 3 evaluates the
role of ED consumption in the relationship between ADHD symptoms and substance use. More
specifically, we examine whether ED consumption mediates the association between baseline
ADHD symptoms among German adolescent substance non-users and substance (tobacco, e-
cigarette, alcohol, and marijuana) use initiation in 12 months. Collectively findings from the
proposed studies contribute to a more comprehensive understanding of ED use behavior and
ADHD and their associated consequences. The following aims and hypotheses will be tested:
Aim 1: To estimate the prevalence of ED use and describe the epidemiology by
demographic characteristics in both adolescent samples. To examine whether ED consumption is
associated with tobacco, e-cigarette, alcohol, and marijuana use. To evaluate whether these
associations differ among U.S. and German adolescents.
Hypothesis 1: Similar demographic (i.e., age, gender), environmental (peer use), and
intrapersonal (risk-taking, depression symptoms) factors will be correlated with ED use in both
samples. ED consumption is associated with tobacco, e-cigarette, alcohol, and marijuana use in
both samples.
Aim 2: To examine whether ED consumption among a German sample of substance non-
users at baseline is associated with tobacco, e-cigarette, alcohol, and marijuana use initiation in
12 months.
Hypothesis 2: ED consumption among a sample of substance non-users is associated
with tobacco, e-cigarette, alcohol, and marijuana use initiation in 12 months. More frequent ED
consumption is longitudinally associated with higher odds of substance use initiation.
22
Aim 3: To examine whether risk-taking moderates the association between ED use and
substance use initiation.
Hypothesis 3: Risk-taking moderates the association between ED use and substance use
initiation. ED users with high risk-taking scores (high-risk adolescents) are more likely to initiate
substance use behavior than ED users with low risk-taking scores (low-risk adolescents).
Aim 4: To examine whether self-reported ADHD symptoms are correlated with more
frequent ED consumption.
Hypothesis 4: Self-reported ADHD symptoms are correlated with more frequent ED
consumption.
Aim 5: To evaluate whether self-reported ADHD symptoms among a sample of ED non-
users at baseline are associated with ED use initiation in 12 months.
Hypothesis 5: Self-reported ADHD symptoms among a sample of baseline ED non-users
are associated with ED use initiation in 12 months.
Aim 6: To examine whether, after controlling for baseline self-reported ADHD
symptoms, ED use at baseline predicts an increase in self-reported ADHD symptoms frequency
in 12 months to evaluate potential reverse causation.
Hypothesis 6: After controlling for baseline self-reported ADHD symptoms, more
frequent ED use at baseline predicts an increase in self-reported ADHD symptoms frequency in
12 months.
Aim 7: To examine whether ED use frequency among baseline substance non-users
mediates the association between self-reported ADHD symptoms and tobacco, e-cigarette,
alcohol, and marijuana use initiation in 12 months.
23
Hypothesis 7: The association between ADHD symptoms and tobacco, e-cigarette,
alcohol, and marijuana use initiation in 12 months is mediated by baseline ED use frequency.
24
Chapter 1: Examining the Association Between Energy Drink Consumption and Tobacco,
Alcohol, And Marijuana Use Among German and U.S. Adolescents.
INTRODUCTION
Energy drinks (ED) are beverages that contain high levels of caffeine (80-500 mg, as
compared with less than 50 mg in a typical can of cola) and are marketed to boost energy,
decrease feelings of tiredness, and enhance mental alertness and concentration (Harris &
Munsell, 2015; Leal & Jackson, 2018; Reissig et al., 2009; Vercammen et al., 2019). Additional
ingredients include amino acids, sugars or sweeteners, guarana, taurine, ginseng, L-carnitine,
herbal supplements, and B vitamins, the short- and long-term effects of which for the most part
are unknown (Leal & Jackson, 2018; McCusker et al., 2006; Reissig et al., 2009; Seifert et al.,
2011; Vercammen et al., 2019).
Aggressive marketing tailored towards youth through carefully crafted campaigns,
including sponsorship of events that appeal to this age group (e.g., snowboarding), and product
placement in video games and social media, resulted in exponential growth of their sales among
minors in recent years (AIM Market Insight, 2015; Emond et al., 2015; Facts, 2013; Harris,
2013; Harris et al., 2011). This rapid expansion of ED sales has raised concerns among health
professionals regarding potential adverse health consequences for children and adolescents
(Schneider & Benjamin, 2011). For instance, the consumption of EDs has been linked to
difficulty breathing, prolonged QT-interval, heart failure, seizures, and sleep disturbances among
adolescents and young adults (Bashir et al., 2016; Ehlers et al., 2019; Hammond et al., 2018;
Park et al., 2016; Seifert et al., 2011).
Previous studies have identified predictors of smoking and alcohol use initiation,
including older age, parental and peer substance use, risk-taking, depression, intention to try
25
tobacco and alcohol, stress, sensation seeking, low attachment to family, low socioeconomic
status (SES), and poor school performance (Covey & Tam, 1990; Fisher et al., 2007; Tyas &
Pederson, 1998). Two prospective studies have shown that ED use frequency at baseline is
associated with higher alcohol use frequency at 12-16-month follow-up (Choi et al., 2016;
Miyake & Marmorstein, 2015); however, these two studies were conducted among current
adolescent substance users and failed to control for a current cigarette, e-cigarette, hookah or
marijuana use, as well as other important confounders. Thus, whether ED consumption is
associated with the risk of initiating tobacco, e-cigarettes, alcohol, and other substances is
unknown. Like other stimulants, ED ingredients (i.e., caffeine, guarana, and taurine) can increase
the release of pleasure reward neurotransmitters (such as dopamine and nor-epinephrine) and
sensitize the mesolimbic reward pathways of the brain (Robinson & Berridge, 1993, 2008),
which potentially may result in substance use and addiction (Leal & Jackson, 2018; Woolsey et
al., 2014). Alternatively, ED use may be merely a marker for risk-taking adolescents, reflecting a
common liability for substance use in general (Jessor, 1991).
If ED use is a risk factor for initiation of tobacco, e-cigarettes, alcohol, and other drugs,
then the public health consequences of ED consumption among minors might be more severe
than initially presumed. Previous studies have emphasized the importance of the cultural and
social environment in understanding and explaining the prevalence and patterns of adolescent
substance use (Bauman & Phongsavan, 1999; Link, 2008). The respective social and cultural
environments can directly or indirectly shape substance and ED use behaviors by altering factors
correlated with these behaviors. This study extends previous work and fills gaps in the literature
by analyzing data from two large (U.S. and German) school-based adolescent samples. The first
objective of this study is to estimate the prevalence of ED use and describe the epidemiology by
26
demographic characteristics in U.S. and German adolescent samples. Second, cross-sectional
associations between ED consumption and substance use (tobacco, e-cigarette, alcohol, and
marijuana) are tested. Third, in the German sample of baseline substance non-users, whether
baseline ED consumption is associated with the risk of initiating tobacco, e-cigarette, alcohol,
and marijuana use in 12 months is evaluated. Finally, whether risk-taking moderates these
associations is examined. It is hypothesized that similar demographic (i.e., age, gender),
environmental (peer use), and intrapersonal (risk-taking, depression symptoms) factors are
correlated with ED use in both samples. Additionally, it is hypothesized that ED consumption
among a sample of substance non-users is associated with tobacco, e-cigarette, alcohol, and
marijuana use initiation in 12 months, while risk-taking moderates these associations. The
heuristic model of this study is summarized in Figure 2.
METHODS
German Sample Description
Data were collected as part of an ongoing longitudinal cohort survey of substance and ED
use among adolescents from six Federal states of Germany: Baden-Württemberg, Mecklenburg-
West-Pomerania, North-Rhine-Westphalia, Rhineland-Palatinate, Saxony, and Schleswig-
Holstein. Each state was randomly selected from one of six Nielsen regions, which cluster areas
with similar purchasing power and consumer behavior (GmbH, 2016). A total of 627 schools
were identified in randomly selected sub-regions within each state; all of them were invited to
participate in the study, while eighty-three schools agreed to participate.
27
Figure 2. Study 1 heuristic model.
Adolescents’ verbal assent and parents’ written informed consent were obtained before
conducting the study. Data were collected by trained research staff or school personnel through
self-completed anonymous questionnaires during one school hour (45 min). To link the baseline
and follow-up questionnaires, students were asked to generate an anonymous seven-digit
individual code, a procedure that had been tested in previous studies, slightly modified for this
study (Galanti et al., 2007; Hanewinkel et al., 2011; Morgenstern et al., 2013). Data analyses
involved two assessment waves that took place approximately 12 months apart: baseline (Fall-
Spring of the school year 2018-2019) and 12-month follow-up (Fall-Spring of the school year
2019-2020).
28
Participation in the study was voluntary, and all participants had the option of
withdrawing from the study at any time without a penalty. After completing the survey, the
questionnaires were placed in an envelope and sealed in front of the class. Students were assured
that their individual information would not be seen by parents or school administrators
(Morgenstern et al., 2013). Study implementation was approved by the ministries of cultural
affairs of the six involved states. Ethical approval was obtained from the Ethical Committee of
the German Psychological Society (Ref. No. RH 042015_1).
U.S. Sample Description
Data were collected as part of an ongoing Trends in Tobacco Use Survey (TITUS). This
school-based longitudinal study was conducted in Southern California. In the Winter-Spring of
the school year 2019-2020, trained research staff members visited each 9
th
-grade classroom to
explain the study, obtain student assent, and provide contact information to obtain parental
consent before conducting a survey.
The survey was administered in the classroom, proctored by a trained data bilingual collector, via
the web through the REDCap platform that had desktop/laptop and mobile/tablet versions to
allow participants flexibility in which device they used to complete the survey. Students were
assured that participation in a study is voluntary and that their responses will be kept confidential
and will be seen only by the researchers, not by their teachers or parents. Students who were not
available for the in-class survey administrations were contacted via email, phone call, or text and
provided a survey link. All students who participated in the study received small gifts (e.g., pens,
stickers) as a token of appreciation, while one of their parents that completed the consent process
(regardless of agree or disagree) received a $5 gift card. This study was approved by the
29
University of Southern California Institutional Review Board. A comparison of study measures
between U.S. and German adolescent samples is reported in Table 1.
German Sample Measures
Energy Drink and Substance Use Behavior. ED and substance use in the German
sample were assessed at each time point (baseline and 12-months follow-up). Lifetime and
current ED use was assessed using the following questions: “Have you ever tried energy drinks
in your life?” (yes or no) and “How often do you currently drink energy drinks?” (not at all, less
than once a month, at least once a month, but not every week, at least once a week, but not every
day, and almost every day). Answers to the latter question other than “Not at all” and “Less than
once a month” were coded as current (past 30-day) use of EDs. Lifetime substance use behaviors
were examined with the questions: “Have you ever tried cigarettes /e-cigarettes /hookah /alcohol
/marijuana in your life?”. There were five response options for each question: “Never,” “Few
puffs”/” Just tried a little bit,” “1 to 19 times”, “20-100 times”, and “More than 100 times”.
Answers other than “Never” (and “Just tried a little bit” for alcohol use questions) were coded as
lifetime use of each corresponding product. To examine the current (past 30-day) use of
substances, the following questions were asked: “How often do you currently use cigarettes /e-
cigarettes /hookah /alcohol /cannabis?”. The response options included: “not at all,” “less than
once a month,” “at least once a month, but not every week,” “at least once a week, but not every
day,” and “almost every day”. Answers other than “Not at all” and “Less than once a month”
were coded as past 30-day use of a corresponding substance.
30
Table 1. Comparison of study measures between U.S. and German adolescent samples.
Study variables German Sample U.S. Sample Measures
Identical?
Validation studies/ Notes
ED use and substance use
ED use
Lifetime and past 30-
day use
Lifetime and past 30-
day use
Yes Slightly different wording
Substance use
Lifetime and past 30-
day use
Lifetime and past 30-
day use
Yes Slightly different wording
Sociodemographic factors Yes
German sample has a slightly
wider age range
Gender Male/Female Male/Female Yes N/A
Age, mean (SD) 9-18 years 13-16 years Yes U.S. sample slightly older
Self-reported SES
1-High SES
10-Low SES
4-point scale No -
Ethnic background Not measured
Hispanic or Latino
(Yes or No); Ethnicity
N/A N/A
Environmental Factors
School type Gymnasium/Other Not measured N/A N/A
Peer substance use Yes or no Yes or no Yes Slightly different wording
Intrapersonal Factors
Risk-taking scale
2-item construct; 1-not
at all, 5-very often
UPPS 4-item index; 1-
strongly disagree, 4-
strongly agree
No -
Depression symptoms scale
SDQ, emotional
problems subscale
CES-D-10 No
Correlation between scales
(Walker et al., 2020)
Self-reported school
performance
1 - somewhat worse,
5- much better
Not measured N/A N/A
31
Covariates. To address possible confounding influences, baseline factors correlated with
ED use or outcome substances (i.e., cigarettes, e-cigarettes, hookah, alcohol, and marijuana) in
previous studies were included as covariates (Azagba et al., 2014; Covey & Tam, 1990; Fisher et
al., 2007; Galimov et al., 2019; Masengo et al., 2020; Terry-McElrath et al., 2014; Tyas &
Pederson, 1998; Visram et al., 2016). Sociodemographic, environmental, and intrapersonal
(within-person/ individual characteristics) covariates were assessed.
Sociodemographic. Self-report sociodemographic covariates were age, gender, and SES,
which was measured (Goodman et al., 2001) with the question, “Please place an ‘X’ on the step
that best represents where you think your family stands on the ladder?” (on a ten-point scale
corresponding to a picture of a ladder, ranging from 0 [low income, the worst jobs, the lowest
education] to 10 [high income, the best jobs, the highest education]).
Environmental Factors. Indicators of the proximal environment included the type of
school participants attended (i.e., gymnasium or other) and peer substance use. The German
school system has several types of secondary schools (i.e., Hauptschule, Realschule, Oberschule,
Gemeinschaftsschule, Gymnasium) that differ with regard to the academic skills of their students
and graduation level (i.e., students typically graduate from school after 10
th
– 13
th
grade,
depending on school type). Gymnasia are the most advanced type of secondary school that
strongly emphasizes academic learning.
Peer substance use was assessed by responses to the question, “How many of your friends
use cigarettes /e-cigarettes or e-hookah/ hookah/ alcohol?” (eleven-point scale, ranging from
0=none to 10=all). Each response was dichotomized considering answers of anything other than
“none” as peer cigarette/ e-cigarette/ hookah/ alcohol use.
32
Intrapersonal Characteristics. Risk-taking was measured using a 2-item index
(Stephenson et al., 2003), averaging responses to the following two questions: “How often do
you do dangerous things to have fun?” and “How often do you do exciting things, even if they
are dangerous?” (each on a 5-point scale ranging from “not at all” to “very often”).
Self-reported school performance (Achenbach, 1991; Sharif & Sargent, 2006) was
assessed by asking, “How would you rate your school performance compared to the classmates
in your class?”. Responses corresponded to the following scores: a response of “much worse”
received a score of 1; “somewhat worse,” 2; “about the same,” 3; “somewhat better,” 4; and
“much better,” 5.
Depression symptoms were assessed using the 5-item emotional problems subscale
(Goodman, 2001; Goodman et al., 2003; Muris et al., 2003) of the strengths and difficulties
questionnaire (SDQ; “I worry a lot,” “I am often unhappy, depressed or tearful,” “I am nervous
in new situations,” “I easily lose confidence,” “ I have many fears, I am easily scared”).
Numerous validation studies have shown that the SDQ is a valid tool for assessing different
behavioral aspects of children and adolescents (including depression). The response options
ranged from 0 = “not true”, 1 = “somewhat true”, and 2 = “certainly true”. The answers to these
five items were further summed (Goodman et al., 2003) in the self-reported depression
symptoms frequency scale (Cronbach’s alpha=0.75).
U.S. Sample Measures
Energy Drink and Substance Use Behavior. To examine lifetime use of EDs and
substances (i.e., cigarettes, e-cigarettes, hookah, cigars, alcohol, and marijuana,) in the U.S.
sample, the following matrix item was asked: “Have you ever used the following substances in
your life?” (yes or no). The products type categories included: “A few puffs of a cigarette”, “A
33
whole cigarette”, “Electronic cigarette with nicotine”, “Electronic cigarette without nicotine or
hash oil”, “Juul or similar device”, “Other electronic vaping device”, “Hookah water pipe”,
“Energy drinks [Red Bull, Monster, Rockstar, and NOS]”, “One full drink of alcohol”,
“Blunts”, “Smoking Marijuana”, “Marijuana or THC foods or drinks” and “Electronic device
to vape THC or hash oil”. Similarly, current (past 30-day) ED consumption and substance use
behaviors were assessed with the following item “In the last 30 days, how many total days have
you used...?”. There were seven response options: “0 days”, “1-2 days”, “3-5 days”, “6-9
days”, “10-19 days”, “20-29 days”, “All 30 days”. Answers of anything other than “0 days”
were coded as past 30-day use of a corresponding substance. The substances comprising multiple
product type categories were combined into a single item (i.e., “A few puffs of a cigarette” and
“A whole cigarette” were combined into cigarette use).
Covariates. Sociodemographic, environmental, and intrapersonal (within-person/
individual characteristics) covariates were assessed. Self-report sociodemographic covariates
were age, gender, ethnicity, and SES, which was assessed with the item, “Think about your
family when you were growing up, from birth to age 16. Would you say your family during that
time was pretty well off financially, about average, or poor?”. Response options included:
“Pretty well off financially”, “About average”, “Poor”, and “It varied”.
Environmental Factors. Indicators of the proximal environment included the peer
substance use, which was assessed with the item “How many of your FIVE closest friends use
these products?” (on a seven-point scale, 1-none, 6-all, 7 – not sure). Each response was
dichotomized considering answers of anything other than “none” as peer substance use, while a
response “not sure” was treated as missing. The products type categories included: “Cigarettes”,
“Electronic cigarette with nicotine”, “Electronic cigarette without nicotine or hash oil”, “Juul”,
34
“Hookah water pipe”, “Smoking Marijuana”, “Marijuana or THC foods or drinks” and
“Electronic device to vape THC or hash oil”. The substances comprising multiple product type
categories were collapsed into a single item.
Intrapersonal Factors. Risk-taking was measured using (UPPS) a 4-item index,
averaging responses to the following items: “I quite enjoy taking risks”, “I welcome new and
exciting experiences and sensations, even if they are a little frightening and unconventional”, “I
would like to learn to fly an airplane”, and “I would enjoy the sensation of skiing very fast down
a high mountain slope” (each on a 4-point scale ranging from 1- “strongly disagree” to 4 -
“strongly agree”).
Depression symptoms were assessed using 10-item Centre for Epidemiological Studies
Depression Scale (CES-D-10; Zhang et al., 2012) that evaluates the frequency of past week
depression symptoms. The response options ranged from 0 – “rarely or none of the time; 0-1
day” to 3 – “most or all of the time; 5-7 days”.
Data Analysis
Participant accrual, sample size, and exclusions from both analytic samples are described
first. Then the prevalence of ED use behavior and descriptive statistics are reported. Bivariate
associations between study variables and lifetime/past 30-day ED use in both analytic samples
are reported next. Pearson’s chi-square tests were used for categorical study variables, while t-
tests were performed for continuous variables.
Multilevel logistic regression models were used to evaluate whether ED use (predictor,
coded as 0=never use, 1=ever use, 2=past 30-day use) is cross-sectionally associated with
tobacco, e-cigarette, alcohol, and marijuana use. The multilevel approach was used to control for
the nesting of students (Level 1) within 83 schools in the German sample (ICC=0.10-0.22,
35
students range per school: 22-502) and 6 schools (Level 2) in the U.S sample (ICC=0.04-0.05,
students range per school: 204-389). Separate models were constructed for each binary outcome
(lifetime and past 30-day use (yes or no) of cigarettes, e-cigarettes, hookah, alcohol, and
marijuana) in each analytic sample. All models were fitted with and without adjustment for all
covariates.
Finally, in the German sample of baseline substance never users (n=3874), multilevel
logistic regression models were used to test whether baseline ED consumption (coded as 0=never
use, 1=ever use, 2=past 30-day use) predicted substance use initiation at 12-month follow-up.
Separate multilevel models were constructed for each binary outcome (i.e., ever use of cigarettes,
e-cigarettes, hookah, alcohol, and marijuana). In a further step of the analysis, ED (coded as
0=never use, 1=ever use) x Risk-Taking (dichotomized as low [0] = below the median, high [1]
= above median) interaction term was added to all models to test whether the risk-taking variable
moderated the association between ED use and substance use initiation. All models were fitted
with and without adjustment for all covariates.
Maximum likelihood estimation was used to account for non-normal distributions and
missing data in both analytic samples. Continuous variables were standardized (mean = 0, SD =
1) to facilitate interpretation. All statistical analyses were conducted using Stata software
(version 15.1; Stata Corp, College Station, Texas, USA). Odds ratios (ORs) with 95% CIs were
reported with statistical significance set at P < .05 (2-tailed). Benjamini-Hochberg multiple
testing corrections were applied to control the false-discovery rate at .05. Additional sensitivity
analyses are summarized below in the Results section.
36
RESULTS
German Study Sample Description
Participant accrual, sample size, and exclusions from the analytic sample are depicted in
Figure 3. Among 16,780 eligible participants, 14,242 (85.0%) provided parental consent and
were present on the day of the baseline survey. After excluding 406 (2.9%) students who did not
complete data on baseline key variables (i.e., baseline ED and substance use), the cross-sectional
analytic sample was 13,836. After excluding 4,265 adolescents who graduated from school after
the 10th grade and 3,952 students lost to follow-up, 5,619 (58.7%) completed both assessments.
After excluding 64 (1.1%) participants who did not complete data on follow-up substance use
and 1,681 (29.9%) students who already used tobacco and alcohol at baseline, the longitudinal
analytic sample was 3,874. Participants’ baseline characteristics are presented in Tables 2 and 3.
Attrition analysis. Baseline students with and without follow-up data did not differ by
baseline ED use or any other covariate except for age and school type. That is, participants
without follow-up data were older (p<0.001) and studied in schools other than gymnasia
(p<0.001).
U.S. Study Sample Description
A total of 2,100 9th grade students (14-year-old) were invited to take the survey, while
1,890 (90%) agreed to participate in the study. After excluding 12 (0.6%) participants who did
not complete data on baseline key variables (i.e., ED and substance use), the analytic sample was
1,878. Participants’ baseline characteristics are presented in Tables 2 and 3.
37
Figure 3. The flow of participants in Study 1.
38
Descriptive analyses
Among the 13,836 students included in the German sample, half were boys (51.0%) with
a mean age of 13.0 years (SD=1.8). About half of the students (44.9%) attended high academic
track schools (gymnasia).
Similarly, among the 1,878 participants in the U.S. sample, half were boys (47.1%) with
a mean age of 14.2 (0.4) years. About half of the students (54.7%) self-identified as
Hispanic/Latino. Additionally, 24.4% were Asian, 7.7% were African American/Black, 29.2%
were White, 19.3% were multiethnic, and 27.3% were of other ethnicities (e.g., Filipino, Indian).
Other sociodemographic, environmental, and intrapersonal characteristics for both adolescent
samples (U.S. and German) are reported in Table 2.
ED and substance use behaviors. About half of the German students (55.9%) reported
using EDs in their lifetime, while only 35.1% of U.S. adolescents reported lifetime ED use. Past
30-day ED use prevalence did not differ significantly across the two adolescent samples (17.6%
and 18.5%, respectively). German adolescents reported a higher prevalence (lifetime and past
30-day) of cigarette, hookah, and alcohol use behaviors compared to U.S. adolescents, while
marijuana use was more prevalent in the U.S. sample (see Table 3). Lifetime e-cigarette use was
higher among German adolescents (19.9% and 18.2%, respectively), while past 30-day e-
cigarette use was higher among U.S. adolescents (9.6% and 3.9%, respectively).
39
Table 2. German and U.S. participant characteristics.
Study variables
a
German Sample
(n = 13 836)
U.S. Sample
(n=1878)
p-value
b
Sociodemographic factors
Gender
- Male 51.0 47.1 0.002
Age, mean (SD) 13.0 (1.8) 14.2 (0.4) <0.001
Self-reported SES, mean (SD) 6.8 (1.5) 3.1 (0.8) N/A
c
Hispanic or Latino
- Yes - 54.7 N/A
Ethnic background (Yes)
- Asian - 24.4
N/A
- Black - 7.7
- White - 29.2
- Multi-ethnic - 19.3
- Other - 27.3
Environmental Factors
School type
- Gymnasia 44.9 -
N/A
- Other 55.1 -
Peer substance use
- Cigarettes 49.1 5.9 <0.001
- E – cigarettes 42.7 39.6 0.02
- Hookah 47.2 5.9 <0.001
- Alcohol 64.1 - N/A
- Marijuana - 34.5 N/A
Intrapersonal Factors
Risk-taking scale, mean (SD) 2.3 (1.1) 2.4 (0.8) N/A
c
Depression symptoms scale,
mean (SD)
2.9 (2.4) 1.0 (0.5) N/A
c
Self-reported school
performance, mean (SD)
3.2 (0.9) - N/A
Abbreviations: SES, socioeconomic status; SD, standard deviation
a
Data are expressed as percent (%) unless otherwise indicated
b
For the difference between U.S. and German sample
c
Cannot be compared across samples due to differences in measures used
40
Table 3. ED consumption and substance use behavior among German and U.S. adolescents.
Study variables
a
German Sample
(n = 13 836)
U.S. Sample
(n=1878)
p-value
b
Energy Drink Use
- Lifetime 55.9 35.1 <0.001
- Past 30-day 17.6 18.5 0.37
Lifetime substance use
- Cigarettes 20.6 5.1 <0.001
- E – cigarettes 19.9 18.2 0.08
- Hookah 19.6 2.4 <0.001
- Alcohol 32.6 19.1 <0.001
- Marijuana 13.1 15.4 0.006
Past 30-day substance use
- Cigarettes 5.8 1.4 <0.001
- E – cigarettes 3.9 9.6 <0.001
- Hookah 4.4 0.7 <0.001
- Alcohol 13.7 7.7 <0.001
- Marijuana 3.8 9.3 <0.001
a
Data are expressed as percent (%)
b
For the difference between U.S. and German samples
41
Cross-sectional analyses
Bivariate analyses. Bivariate comparisons demonstrated that compared to non-users,
lifetime and past 30-day ED users (both U.S. and German) tended to be older, were more likely
to be males, were more likely to report lifetime and current substance use, had higher risk-taking
scores, reported more depression symptoms, and had more peers using substances (all p-
values<0.05). Additionally, in the U.S. sample, those identified as Hispanic/Latino, White, and
multiethnic were more likely to use EDs, while those identified as Asian were less likely to
report ED use (p<0.05). Detailed information on bivariate associations between study variables
and lifetime/current ED consumption is reported in Tables 4 and 5.
Cross-sectional multilevel models. Multilevel logistic regression models revealed that
after adjustment for sociodemographic, environmental, and intrapersonal covariates, ED
consumption was associated with higher odds of a lifetime as well as past 30-day substance use
(cigarettes, e-cigarettes, hookah, alcohol, and marijuana) in the German sample and with lifetime
substance use (except hookah use) and past 30-day e-cigarette/alcohol/marijuana use in the U.S.
sample (see Table 6).
For instance, multilevel logistic regression model demonstrated that after covariate
adjustment, German lifetime ED users (Adjusted Odds Ratio [AOR] 4.45 [95% CI, 3.63-5.46])
as well as past 30-day ED users (AOR 7.94 [95% CI, 6.37-9.89]) had higher odds of using
cigarettes in their lifetime compared to ED non-users. Similarly, compared to non-users, lifetime
and past 30-day ED users in the U.S. sample had higher odds of using cigarettes in their lifetime
(AOR 4.80 [95% CI, 1.42-16.18] for lifetime ED use versus non-use and AOR 6.51 [95% CI,
1.99-21.33] for past 30-day ED use versus non-use). The results of all cross-sectional multilevel
logistic regression models are reported in Table 6.
42
Table 4. German and U.S. participant characteristics separated by lifetime ED use.
Study variables
a
German Sample U.S. Sample
Lifetime ED use
p-value
b
Lifetime ED use
p-value
b
Yes No Yes No
Sociodemographic factors
Male 56.5 44.0 <0.001 50.3 45.4 0.04
Age, mean (SD) 13.7 (1.6) 12.2 (1.6) <0.001 14.3 (0.5) 14.2 (0.4) 0.004
Self-reported SES, mean
(SD)
6.7 (1.6) 6.9 (1.4) <0.001 3.0 (0.8) 3.1 (0.7) <0.001
Hispanic or Latino (Yes) - - - 61.2 51.2 <0.001
Ethnic background (Yes)
- Asian - - - 18.0 27.9 <0.001
- Black - - - 8.2 7.4 0.54
- White - - - 33.3 26.9 0.004
- Multi-ethnic - - - 23.8 16.9 <0.001
- Other - - - 28.0 26.3 0.61
Environmental Factors
School type (Gymnasia) 36.1 56.2 <0.001 - - -
Peer substance use
- Cigarettes 61.3 22.7 <0.001 8.0 4.8 0.008
- E – cigarettes 55.1 15.8 <0.001 56.0 30.5 <0.001
- Hookah 59.8 19.9 <0.001 10.7 3.3 <0.001
- Alcohol 75.0 40.4 <0.001 - - -
- Marijuana - - - 55.2 23.2 <0.001
Intrapersonal Factors
Risk-taking scale, mean (SD) 2.6 (1.1) 2.0 (0.9) <0.001 2.6 (0.8) 2.3 (0.7) <0.001
Depression symptoms scale,
mean (SD)
3.1 (2.4) 2.6 (2.2) <0.001 1.2 (0.6) 1.0 (0.5) <0.001
Self-reported school
performance, mean (SD)
3.1 (0.9) 3.3 (0.8) <0.001 - - -
Lifetime substance use
- Cigarettes 34.7 2.8 <0.001 9.9 2.5 <0.001
- E – cigarettes 34.4 1.6 <0.001 36.5 8.2 <0.001
- Hookah 33.2 2.3 <0.001 3.9 1.6 0.001
- Alcohol 50.5 10.0 <0.001 36.8 9.4 <0.001
- Marijuana 18.3 1.8 <0.001 32.6 6.2 <0.001
Past 30-day substance use
- Cigarettes 10.0 0.5 <0.001 3.2 0.5 <0.001
- E – cigarettes 6.8 0.3 <0.001 22.7 3.2 <0.001
- Hookah 7.6 0.4 <0.001 1.8 0.1 <0.001
- Alcohol 22.6 2.4 <0.001 16.8 2.8 <0.001
- Marijuana 5.4 0.6 <0.001 23.0 2.6 <0.001
Abbreviations: SES, socioeconomic status; SD, standard deviation
a
Data are expressed as percent (%) unless otherwise indicated
b
For the difference between lifetime ED users and non-users
43
Table 5. German and U.S. participant characteristics separated by past 30-day ED use.
Study variables
a
German Sample U.S. Sample
Past 30-day ED use p-
value
b
Past 30-day ED use
p-value
b
Yes No Yes No
Sociodemographic factors
Male 62.5 48.5 <0.001 51.0 46.2 0.11
Age, mean (SD) 14.0 (1.5) 12.8 (1.8) <0.001 14.3 (0.5) 14.2 (0.4) 0.003
Self-reported SES, mean
(SD)
6.7 (1.7) 6.8 (1.5) <0.001 3.0 (0.8) 3.1 (0.8) 0.003
Hispanic or Latino (Yes) - - - 65.2 52.3 <0.001
Ethnic background (Yes)
- Asian - - - 16.1 26.3 <0.001
- Black - - - 6.1 8.0 0.21
- White - - - 34.3 28.0 0.02
- Multi-ethnic - - - 24.8 18.1 0.004
- Other - - - 27.1 27.4 0.92
Environmental Factors
School type (Gymnasia) 27.0 48.8 <0.001 - - -
Peer substance use
- Cigarettes 75.1 41.5 <0.001 7.4 5.6 0.21
- E – cigarettes 72.2 34.0 <0.001 62.0 34.5 <0.001
- Hookah 75.1 39.0 <0.001 14.6 4.0 <0.001
- Alcohol 83.7 58.3 <0.001 - - -
- Marijuana - - - 61.2 28.5 <0.001
Intrapersonal Factors
Risk-taking scale, mean (SD) 3.0 (1.1) 2.2 (1.0) <0.001 2.7 (0.8) 2.3 (0.7) <0.001
Depression symptoms scale,
mean (SD)
3.3 (2.6) 2.8 (2.3) <0.001 1.2 (0.5) 1.0 (0.5) <0.001
Self-reported school
performance, mean (SD)
3.0 (1.0) 3.2 (0.8) <0.001 - - -
Lifetime substance use
- Cigarettes 51.8 14.0 <0.001 11.8 3.5 <0.001
- E – cigarettes 53.6 12.7 <0.001 40.9 13.0 <0.001
- Hookah 50.1 13.0 <0.001 4.6 1.9 0.003
- Alcohol 63.0 26.1 <0.001 41.5 14.0 <0.001
- Marijuana 26.2 9.3 <0.001 36.9 10.6 <0.001
Past 30-day substance use
- Cigarettes 18.6 3.1 <0.001 5.2 0.6 <0.001
- E – cigarettes 14.7 1.6 <0.001 29.5 5.4 <0.001
- Hookah 15.6 2.0 <0.001 2.9 0.2 <0.001
- Alcohol 32.8 9.6 <0.001 24.5 3.9 <0.001
- Marijuana 8.1 2.6 <0.001 29.9 4.9 <0.001
Abbreviations: SES, socioeconomic status; SD, standard deviation
a
Data are expressed as percent (%) unless otherwise indicated
b
For the difference between past 30-day ED users and non-users
44
Table 6. Cross-sectional associations between ED consumption and substance use behavior among German and U.S. adolescents
Lifetime use
Cigarettes E-cigarettes Hookah Alcohol Marijuana
AOR (95% CI) AOR (95% CI) AOR (95% CI) AOR (95% CI) AOR (95% CI)
German Sample (n=13 836)
ED use (never use=ref)
- Lifetime use
a
4.45 (3.63-5.46)
*
7.01 (5.45-9.02)
*
4.52 (3.62-5.63)
*
3.41 (2.96-3.94)
*
5.14 (3.78-6.98)
*
- Past 30-day use 7.94 (6.37-9.89)
*
12.72 (9.77-16.55)
*
7.46 (5.91-9.43)
*
4.97 (4.17-5.92)
*
7.33 (5.33-10.07)*
U.S. Sample (n=1878)
ED use (never use=ref)
- Lifetime use
a
4.80 (1.42-16.18)
*
3.79 (1.97-7.29)
*
4.84 (0.95-24.74) 3.51 (1.93-6.40)
*
3.99 (1.93-8.25)
*
- Past 30-day use
6.51 (1.99-21.33)
*
5.83 (3.12-10.90)
*
1.14 (0.17-7.79) 6.48 (3.74-11.24)
*
3.54 (1.77-7.07)
*
Past 30-day use
Cigarettes E-cigarettes Hookah Alcohol Marijuana
AOR (95% CI) AOR (95% CI) AOR (95% CI) AOR (95% CI) AOR (95% CI)
German Sample (n=13 836)
ED use (never use=ref)
- Lifetime use
a
2.60 (1.71-3.96)
*
2.64 (1.50-4.66)
*
2.85 (1.67-4.88)
*
3.17 (2.55-3.93)
*
3.46 (2.04-5.88)
*
- Past 30-day use 5.07 (3.32-7.75)
*
6.31 (3.59-11.08)
*
6.90 (4.05-11.77)
*
4.89 (3.87-6.16)
*
4.08 (2.37-7.01)
*
U.S. Sample (n=1878)
ED use (never use=ref)
- Lifetime use
a
- 2.39 (0.90-6.40) - 0.96 (0.29-3.20) 4.87 (1.73-13.68)
*
- Past 30-day use - 8.39 (3.55-19.83)
*
- 8.42 (3.91-18.12)
*
7.81 (3.11-19.62)
*
Abbreviations: SES, socioeconomic status; AOR, Adjusted Odds Ratio; CI, Confidence Interval; ref, reference
All models are adjusted for sociodemographic, environmental, and intrapersonal covariates.
* Statistically significant after Benjamini-Hochberg corrections for multiple testing to control the false-discovery rate at .05 (based on 2-tailed corrected P-value).
a
This category includes lifetime ED users but excludes past 30-day ED users
45
Longitudinal Analyses
In the sample of German adolescents who never tried substances at baseline (n=3874),
6.4% of participants had initiated cigarette smoking, 6.0% e-cigarette use, 5.5% hookah use,
18.3% alcohol use, and 2.7% marijuana use in the subsequent 12 months. The results of
unadjusted multilevel logistic regression models demonstrated that (see Table 7) Baseline ED
users (lifetime and past 30-day) who never used substances at baseline had significantly higher
odds of initiating cigarette (OR 4.82 [95% CI, 3.59-6.47] for lifetime ED use versus non-use and
OR 8.96 [95% CI, 5.77-13.90] for past 30-day ED use versus non-use), e-cigarette (6.30 [95%
CI, 4.61-8.61] and 9.45 [95% CI, 5.93-15.06]), hookah (5.92 [95% CI, 4.29-8.17] and 8.39 [95%
CI, 5.15-13.68]), alcohol (3.18 [95% CI, 2.64-3.82] and 4.86 [95% CI, 3.44-6.88]) and marijuana
use (3.94 [95% CI, 2.34-6.63] and 8.70 [95% CI, 4.07-18.56]) in the subsequent 12 months
compared to ED non-users
After adjustment for sociodemographic, environmental, and intrapersonal covariates, the
magnitudes of these associations reduced but remained significant in the adjusted models and
after applying the Benjamini-Hochberg correction for multiple comparisons (see Table 7). For
instance, after covariate adjustment lifetime (AOR 1.92 [95% CI, 1.50-2.45]) and past 30-day
ED users (AOR 2.79 [95% CI, 1.74-4.49]) had significantly higher odds of initiating alcohol use
in the subsequent 12 months compared to ED non-users.
Moderation models. Moderation analyses were conducted for each of the five dependent
variables of interest (i.e., cigarette, e-cigarette, hookah, alcohol, and marijuana). The results of
moderation models demonstrated a statistically significant interaction between ED use and risk-
taking (AOR 10.66 [95% CI 2.17-52.39]) in the model predicting marijuana use initiation. No
other moderation models were significant. In a subgroup analysis stratified by risk-taking
46
categories, lifetime ED use was significantly associated with marijuana use initiation (AOR
14.47 [95% CI, 3.38-62.05]) among high risk-taking adolescent subgroup, but not in the low
risk-taking subgroup (p=0.23). Those in the high-risk taking subgroup had generally higher
marijuana use initiation rates at 12 months follow-up than those in the low risk-taking subgroup;
these rates were also higher among lifetime ED users compared to non-users (see Figure 4).
Figure 4. Association between lifetime ED use at baseline and marijuana use initiation at follow-
up stratified by risk-taking at baseline.
Footnotes: ED - energy drink; RT- risk-taking
1.7%
0.5%
2.9%
8.0%
0%
4%
8%
12%
Low RT Hight RT
Marijunana use initiation rate
ED non-users ED users
47
Supplementary Analyses. In parallel (sensitivity) analyses missing data were handled
using multivariate imputations by chained equations (MICE) method for missing at random
assumptions and the available covariate data, 25 imputed data sets were created. The pooling of
the regression estimates followed Rubin’s rule (White et al., 2011). All significant differences in
these analyses remained and were in the same direction.
An additional analysis was conducted to examine the evidence for the reversed temporal
ordering (i.e., cigarette, e-cigarette, hookah, alcohol, and marijuana use predicting ED initiation).
These analyses included substance users at baseline but excluded lifetime ED users to predict
initiation of ED use. In unadjusted and adjusted models, baseline cigarette, e-cigarette, hookah,
alcohol, and marijuana use did not predict ED initiation in 12 months.
48
Table 7. Predictors of initiating substance use at 12-month follow-up among baseline substance never users in the German
longitudinal sample.
Cigarettes E-cigarettes Hookah Alcohol Marijuana
OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
Unadjusted Models
ED use (never use=ref)
- Lifetime use
b
4.82 (3.59-6.47)
*
6.30 (4.61-8.61)
*
5.92 (4.29-8.17)
*
3.18 (2.64-3.82)
*
3.94 (2.34-6.63)
*
- Past 30-day use 8.96 (5.77-13.90)
*
9.45 (5.93-15.06)
*
8.39 (5.15-13.68)
*
4.86 (3.44-6.88)
*
8.70 (4.07-18.56)
*
Adjusted Models
ED use (never use=ref)
- Lifetime use
b
3.65 (2.40-5.55)
*
3.39 (2.20-5.21)
*
3.56 (2.25-5.65)
*
1.92 (1.50-2.45)
*
2.71 (1.54-4.75)
*
- Past 30-day use 5.72 (3.05-10.73)
*
4.15 (2.18-7.89)
*
4.12 (2.09-8.14)
*
2.79 (1.74-4.49)
*
4.24 (1.75-10.26)
*
Gender (female=ref) 1.28 (0.86-1.91) 1.42 (0.95-2.13) 1.37 (0.89-2.11) 1.04 (0.81-1.33)
*
2.49 (1.33-4.67)
*
School type (other=ref) 0.74 (0.47-1.16) 0.65 (0.42-1.00) 0.71 (0.42-1.20)
*
1.44 (1.04-1.98)
*
1.85 (0.98-3.50)
Age
a
1.29 (0.91-1.82) 1.11 (0.77-1.58) 1.13 (0.77-1.66) 1.91 (1.52-2.40) 2.15 (1.43-3.24)
*
Self-report SES
a
0.95 (0.79-1.15) 1.24 (1.02-1.51) 1.00 (0.82-1.22) 1.17 (1.03-1.33) 1.11 (0.84-1.46)
School performance
a
1.04 (0.86-1.25) 0.84 (0.69-1.01) 0.77 (0.63-0.94) 1.07 (0.95-1.20) 0.94 (0.73-1.23)
Risk-taking
a
1.22 (0.99-1.49) 1.18 (0.96-1.45) 1.15 (0.92-1.43) 1.36 (1.18-1.56)
*
1.37 (1.04-1.81)
Depression symptoms
a
1.01 (0.83-1.24) 1.02 (0.83-1.25) 0.95 (0.76-1.19) 1.02 (0.89-1.16) 0.85 (0.62-1.16)
Peer use (none=ref) 1.40 (0.93-2.10) 2.93 (1.94-4.42)
*
2.70 (1.77-4.13)
*
2.51 (1.98-3.18)
*
-
Moderation Models
ED use (never use=ref)
- Lifetime use 3.79 (2.33-6.16)
*
3.16 (1.90-5.25)
*
3.87 (2.21-6.79)
*
2.08 (1.56-2.77)
*
1.45 (0.73-2.88)
ED use x Risk-taking 1.06 (0.45-2.49) 1.38 (0.56-3.36) 0.91 (0.37-2.22) 0.88 (0.55-1.43) 10.66 (2.17-52.39)
*
Abbreviations: SES, socioeconomic status; OR, Odds Ratio; CI, Confidence Interval; ref, reference
* Statistically significant after Benjamini-Hochberg corrections for multiple testing to control the false-discovery rate at .05 (based on 2-tailed corrected P-value).
a
Rescaled (mean = 0, SD = 1) such that the ORs indicate the change in odds in the outcome associated with an increase in 1 SD unit on the continuous covariate scale.
b
This category includes lifetime ED users but excludes past 30-day ED users
49
DISCUSSION
This study estimated the prevalence of ED use and evaluated cross-sectional and
longitudinal associations between ED use and substance use behaviors by analyzing data from
two large (U.S. and German) school-based adolescent samples. Current ED use was reported by
roughly 20% of the participants in both adolescent samples. These prevalence rates are consistent
with those observed in U.S., Canadian and Australian national surveys and those reported by
European Food Safety (Azagba et al., 2014; Miller et al., 2018; Trapp et al., 2020; Zucconi et al.,
2013). ED use was cross-sectionally associated with substance use variables (i.e., cigarettes, e-
cigarettes, hookah, alcohol, and marijuana) in both adolescent samples, even after controlling for
covariates. Moreover, a longitudinal analysis conducted among German adolescents
demonstrated that ED use was prospectively associated with increased risk of substance use
initiation. Associations were consistent across unadjusted and adjusted models, multiple
substance use outcomes (i.e., cigarettes, e-cigarettes, hookah, alcohol, and marijuana), and
various sensitivity analyses. Collectively, these results support the hypothesis that ED may serve
as the gateway drug to other substances.
Current ED use prevalence was roughly identical across two samples (17.6% vs. 18.5%),
while lifetime ED use prevalence was slightly higher among German adolescents (55.9% vs.
35.1%). Previous studies demonstrated that many adolescents prefer to use EDs mixed with
alcohol (McKetin et al., 2015). Hence, higher lifetime ED use prevalence among German
adolescents may be partially explained by the fact that lifetime alcohol use prevalence was also
higher among German adolescents (32.6% vs. 19.1) compared to the U.S. adolescents (which can
be attributed to the lower minimum legal drinking age in Germany [16 years for undistilled
alcoholic beverages and 18 years for spirits] compared to U.S. [21 years] (Centers for Disease
50
Control and Prevention, 2020; European Union Agency for Fundamental Rights, 2018)). Similar
demographic, environmental, and intrapersonal covariates were associated with ED use behavior
in both adolescent samples. For instance, boys and older adolescents were more likely to
consume EDs. Additionally, risk-taking, more frequent depression symptoms, substance use, and
peer substance use were associated with ED use behavior. These results are consistent with
previous studies conducted among children and adolescents (Azagba et al., 2014; Galimov et al.,
2019; Gallimberti et al., 2013; Kaur et al.; Leal & Jackson, 2018; Masengo et al., 2020;
Sampasa-Kanyinga et al., 2020; Visram et al., 2016). Interestingly, those identified as
Hispanic/Latino and multiethnic were more likely to use EDs, while those identified as Asian
were less likely to use EDs. Indeed, additional nationwide studies involving large youth samples
are needed to better understand the reason for variation in ED use patterns between different
ethnic groups.
In line with the study hypothesis and consistent with the findings from previous cross-
sectional studies, ED consumption was associated with lifetime and current substance use in both
adolescent samples. (Azagba et al., 2014; Galimov et al., 2019; Hamilton et al., 2013; Miller,
2008a; Terry-McElrath et al., 2014). Additionally, consistent with the second study hypothesis,
ED consumption at baseline among German adolescents who never used substances in their
lifetime was prospectively associated with cigarette, e-cigarette, hookah, alcohol, and marijuana
use initiation in subsequent 12 months. Several mechanisms may explain the observed
association between ED use and substance use initiation. First, the effects of ED consumption on
substance use behaviors may be interpreted with a social development context (Li & Lerner,
2011). Children and adolescents spend significant amounts of time in school, which is often
accompanied by new academic demands and social influences such as peer modeling (Benner,
51
2011). The findings of this study have demonstrated that ED use is positively associated with
depression symptoms and negatively associated with academic performance. Thus, the effects of
EDs may become a means of coping with increasing and stressful demands in adolescents’ lives.
Furthermore, adolescence is characterized by uneven cognitive brain development. Neural
structures underlying motivation to seek out novel experiences develop more rapidly than those
involving self-regulation and effective decision making (Steinberg, 2008). Consequently, risky
behaviors, such as smoking, alcohol, and illicit drug use, increase during adolescence.
Second, common risk factors may be responsible for the use of ED and other substances.
For instance, some adolescents may have high sensation seeking-traits or experience a low level
of parental monitoring, making them more prone to try new substances (Fisher et al., 2007; Tyas
& Pederson, 1998; Wills et al., 1994). Additionally, other teens may be more likely to use EDs
prior to other substances because of beliefs that EDs are a type of sports drink and not harmful
(Kumar et al., 2015), youth-targeted marketing (Emond et al., 2015; Hammond & Reid, 2018;
Harris, 2013; Stacey et al., 2017), ease of access, and wide availability of EDs due to absence of
the laws restricting sales to minors (Costa et al., 2014; Harris & Munsell, 2015; Visram et al.,
2016; Visram et al., 2017). To address the possible influence of common risk factors, all
multilevel models were adjusted for sociodemographic, environmental, and intrapersonal
characteristics that could potentially overlap with the use of both types of products. After
adjustment for these factors, the OR estimates associated with ED use reduced but remained
significant.
In line with the Problem Behavior Theory (Donovan et al., 1991; Jessor, 1991; Jessor et
al., 2003), it is also possible that ED use is merely a marker for risk-taking adolescents, reflecting
a common liability for substance use in general. Nonetheless, contrary to the third study
52
hypothesis, risk-taking did not moderate the association between baseline ED use and subsequent
substance use initiation, except for the model predicting marijuana use initiation. In fact, baseline
ED use predicated marijuana use initiation among high risk-taking adolescent subgroup, but not
in the low risk-taking subgroup. This finding is consistent with the results of the previous
longitudinal studies indicating that deviant behaviors and risk preference are associated with
increased marijuana use among adolescents (Colder et al., 2013; Hampson et al., 2008; Moss et
al., 2019). One may speculate that there is a progressive and hierarchical sequence of stages of
drug use that begins with EDs in adolescence and proceeds to legal substances, such as
cigarettes, e-cigarettes, hookah, and alcohol, and further to marijuana use, which then leads to
other illicit drugs use in adulthood (Kandel, 1975; Kandel et al., 1992; Nkansah-Amankra, 2020;
Nkansah-Amankra & Minelli, 2016; Vanyukov et al., 2012; Wagner & Anthony, 2002). It is
possible that high-risk adolescents can “skip the legal substances stage” and progress directly
from EDs to marijuana use.
Finally, one can argue that the observed link between ED consumption and substance use
may be explained by cross-sensitization theory. According to this theory, repeated and excessive
consumption of one specific substance enhances the response to that substance (i.e., it makes the
substance more rewarding) as well as to other drugs that act at the same neurobiological sites,
which may produce hyper-responsive reactions to the other substances (Hellberg et al., 2019;
Robinson et al., 1985; Sussman, 2017; Temple, 2009). Consequently, consumption of one drug
can act as a “gateway” to the use of other drugs. Neurobiological research suggests that caffeine
consumption can stimulate striatal pathways that are associated with the production of dopamine
and reward sensitivity (Retzbach et al., 2014). This area of the brain is particularly responsive
during early adolescence (Galván, 2010). Thus, it is possible that repeated ED consumption can
53
sensitize adolescents to the rewarding effects of subsequent use of other substances. Supporting
this hypothesis, although it remains possible that confounders not measured in this study may
explain the association between ED use and initiation of substance use, it is also plausible that
exposure to EDs may play a role in the risk of substance use initiation. Experimental (animal)
studies utilizing a sequential paradigm are warranted to test the causal validity of this gateway
hypothesis.
Limitations
Several limitations should be taken into account. The U.S. sample was limited to students
studying in the schools located in Southern California, which might limit generalizability to other
geographic areas. To enhance generalizability, future research should include students from the
schools across different regions in the U.S. Additionally, the U.S. sample was limited to cross-
sectional data; thus, causality in the relationships seen here cannot be inferred. Future research
should use longitudinal data to move beyond the associational findings of this study. Some of the
predictors used in both samples are single-item measures; hence, future studies should include
multiple items to assess the construct of interest to reduce measurement error and increase
reliability. Moreover, some of the measures differed across two samples, which might have
limited comparability between the two studies. Loss to follow-up could have affected
generalizability (German sample). Older participants and those who studied in schools other than
gymnasia were more likely to drop out from the study; thus, the results of the study may have
limited generalizability to high-risk adolescents. While this study focused on ED use, overall as
well as exact amounts of caffeine consumption were not measured. Given that EDs may not be
the major caffeine source for children and adolescents (Verster & Koenig, 2018), further
research should account for other sources of caffeine to rule out possible confounding. Finally,
54
given the nature of the data (self-reported), recall, social desirability, and cognitive biases may
have affected the results in both adolescent samples. Some other important covariates (e.g.,
advertising exposure and family environment) were not assessed and should be included in
future work.
Conclusions
This study demonstrated a temporal link between ED use and subsequent substance use
initiation. These results highlight an urgent need for policy regulation and restriction of ED
consumption among children and adolescents. It is also crucial for parents, school officials, and
healthcare providers to be aware of signs of excessive EDs consumption and limit their use by
adolescents.
55
Chapter 2: Examining the Temporal Associations Between ADHD Symptoms and Energy
Drink Consumption Among German Adolescents.
INTRODUCTION
Energy drinks (EDs) have become alarmingly popular among adolescents in recent years,
and their sales have grown exponentially (Harris & Munsell, 2015). In 2018, their worldwide
market size was valued at $53 billion and is expected to top $86 billion by 2026 (Roy &
Deshmukh, 2019). EDs contain a high concentration of caffeine (80-500 mg per can) that is
substantially higher than in a standard cup of coffee or a soft drink (Harris & Munsell, 2015;
Reissig et al., 2009). Other ingredients contained in this unique beverage include guarana,
taurine, ginseng, L-carnitine, herbal supplement, amino acids, sugars or sweeteners, and B group
vitamins (Reissig et al., 2009; Seifert et al., 2011).
Initially, ED companies targeted athletes offering improvement in energy and stamina.
However, recently the market has become more youth-oriented, while ED aggressive advertising
appears designed to reach underaged adolescents (Harris, 2013; Harris & Munsell, 2015). Not
surprisingly, a recent study reported that the prevalence of ED consumption among adolescents
increased significantly in recent years (Vercammen et al., 2019). It was reported that 14.7% of
U.S. adolescents consume EDs at least once a week (Larson et al., 2014), while past 30-day ED
use prevalence among Canadian, Australian, and German adolescents ranges from 20 to 28
percent (Azagba et al., 2014; Costa et al., 2016; Galimov et al., 2019). This rapid inflation of ED
sales and increased popularity among minors have raised concerns among public health and
healthcare professionals regarding potential harm for children and adolescents (Schneider &
Benjamin, 2011). In such a way, a few studies have linked ED use with jitteriness, heart failure,
56
gastrointestinal symptoms, sleeping problems, fatigue, and headache (Ali et al., 2015; Gunja &
Brown, 2012; Koivusilta et al., 2016; Seifert et al., 2011).
Additionally, previous studies identified demographic and psychosocial correlates and
predictors of ED consumption among adolescents. It has been shown that males and older
adolescents are more likely to use EDs (Azagba et al., 2014; Emond et al., 2014; Reid et al.,
2017). School factors have also been predictive of ED consumption. For instance, better
academic performance was negatively associated, while school stress was positively associated
with ED use (Galimov et al., 2019). Evidence suggests that ED consumption is also linked with
depression, sensation seeking, poor eating habits, and elevated BMI (Azagba et al., 2014; Emond
et al., 2014; Galimov et al., 2019; Masengo et al., 2020; Poulos & Pasch, 2015). Peer ED
consumption, ED advertising exposure, and curiosity trying EDs are the other factors associated
with EDs use (Galimov et al., 2019). There is evidence that ED use is correlated with a set of
risky behaviors. For instance, a few cross-sectional studies have shown that ED use is associated
with smoking, alcohol use, sexual risk-taking, and violence (Azagba et al., 2014; Emond et al.,
2014; Gallimberti et al., 2013; Hamilton et al., 2013; Miller, 2008a; Terry-McElrath et al., 2014).
Nevertheless, none of the substances (cigarettes, e-cigarettes, hookah, and alcohol) predicted ED
use initiation (Galimov et al., 2019).
It is also possible that ED use is associated with attention-deficit/hyperactivity disorder
(ADHD). ADHD is a neurodevelopmental condition associated with persistent hyperactivity and
impulsivity symptoms and difficulty sustaining attention (American Psychiatric Association,
2013). ADHD is one of the most common psychiatric disorders in children and adolescents. A
recent systematic review that included 179 ADHD prevalence estimates reported an overall
pooled prevalence of 7.2% (95% CI: 6.7 -7.8) among youth (Thomas et al., 2015). It is believed
57
that dysfunctions in mesocortical brain networks and the mesolimbic dopaminergic system are
the primary causes of ADHD (Biederman, 2005; Swanson et al., 2007). There is evidence
suggesting that almost all drugs of abuse increase dopamine activity (Brody et al., 2009; Krause
et al., 2002; Sussman, 2017; Sussman & Ames, 2008), hence, consistent with the self-medication
hypothesis (Krause et al., 2002; Lawrence et al., 2002; Levin et al., 2006), adolescents with
ADHD may choose to self-medicate with nicotine, alcohol, and other stimulating substances to
manage their symptoms. In this line, a few studies suggested that adolescents with ADHD are
more likely to experience an early onset of substance use disorders than non-ADHD individuals
(Biederman et al., 1997; Katusic et al., 2005; Wilens et al., 2011). While the majority of these
studies have been conducted among clinical adolescent samples, recent studies suggested that
self-reported ADHD symptoms among adolescents with no formal ADHD diagnosis are also
associated with substance (alcohol, nicotine, marijuana, cocaine) use behaviors (De Alwis et al.,
2014; Kessler et al., 2012).
Further, evidence suggests that, like psychostimulants, caffeine can trigger mesolimbic
reward pathways, resulting in a release of dopamine (Nall et al., 2016; Nehlig et al., 1992;
Solinas et al., 2002). For that reason, in line with the self-medication hypothesis, adolescents
(similarly to other substances) may self-medicate with ED drinks (which contain high levels of
caffeine, along with other stimulants) to manage their ADHD symptoms (Krause et al., 2002;
Lawrence et al., 2002; Levin et al., 2006). Thus, it is plausible that having ADHD symptoms
among adolescent ED non-users is associated with future risk of ED use initiation. Nonetheless,
current evidence is limited to one cross-sectional study that discovered an association between
ED consumption and self-reported hyperactivity/inattention symptoms and one longitudinal
study conducted among children (Marmorstein, 2016; Schwartz et al., 2015). Although the
58
longitudinal study demonstrated bi-directional associations between ADHD-like symptoms and
ED intake, these findings were limited by the study's small sample size (n=134), limited
covariates list, and children population (mean age = 11.9). This highlights the need for further
prospective studies that would incorporate a representative community sample, older age groups,
and control for important covariates associated with ADHD and ED use.
This study extends previous work and fills gaps in the literature by analyzing data from a
large German school-based adolescent sample. The first aim of this study is to investigate
whether the baseline self-reported ADHD symptoms among adolescents who never used EDs in
their lifetime are predictive of ED use initiation in 12 months. Evidence suggests that the average
age of initial ED consumption ranges between 10-15 years (Costa et al., 2016; Reid et al., 2015;
Trapp et al., 2020). Many of these adolescents have already tried cigarettes, e-cigarettes, and
alcohol or have older peers with access to these substances (Bonnie et al., 2015; Nuyts et al.,
2020; Schneider et al., 2016; Wagenaar & Toomey, 2002). These and other unmeasured factors
(such as ED marketing) may serve as potential confounders in the relationship between self-
reported ADHD symptoms and ED use initiation. Additionally, serval longitudinal studies have
demonstrated that adolescents with ADHD are more likely to experience an early onset of
substance use disorders than non-ADHD individuals (Biederman et al., 1997; Katusic et al.,
2005; Wilens et al., 2011). These substances can cause alterations in dopamine function transfer,
which in turn may give rise to many symptoms of ADHD (Tripp & Wickens, 2009). Hence, it is
not clear whether ADHD self-reports capture the true symptoms of ADHD disorder or whether
these symptoms are simply markers of substance use. To rule out these potential confounding
influences, the model investigating the link between self-reported ADHD symptoms and ED use
59
initiation was also replicated (sensitivity analysis) among a subgroup of children aged 9-10
years.
The second aim of this study is to test the temporal ordering to evaluate the evidence for
potential reverse causation. It is hypothesized that baseline ADHD symptoms among ED non-
users are positively associated with ED initiation in 12 months; while after controlling for
baseline self-reported ADHD symptoms, ED use at baseline predicts an increase in self-reported
ADHD symptoms frequency in 12 months. The heuristic model of this study is summarized in
Figure 5.
Figure 5. Study 2 heuristic model.
60
METHODS
Participants and procedures
Data were collected as part of an ongoing longitudinal cohort survey of substance and ED
use among adolescents from six Federal states of Germany: Baden-Württemberg, Mecklenburg-
West-Pomerania, North-Rhine-Westphalia, Rhineland-Palatinate, Saxony, and Schleswig-
Holstein. Each state was randomly selected from one of six Nielsen areas, which cluster areas
with similar purchasing power and consumer behavior (GmbH, 2016). The German school
system has several types of secondary schools (i.e., Hauptschule, Realschule, Oberschule,
Gemeinschaftsschule, Gymnasium) that mainly differ with regard to the academic skills of their
students and graduation level (i.e., students normally graduate from school after 10
th
– 13
th
grade
[at the age of 17-19], depending on school type). A total of 627 schools were identified in
randomly selected sub-regions within each state; all of them were invited to participate in a
study, while eighty-three schools agreed to participate (Hansen et al., 2018).
Data were collected by trained research staff or school personnel through self-completed
anonymous questionnaires during one school hour (45 min). To link the baseline and follow-up
questionnaires, students were asked to generate an anonymous seven-digit individual code, a
procedure that had been tested in previous studies, slightly modified for this study (Galanti et al.,
2007; Hanewinkel et al., 2011; Morgenstern et al., 2013). Participation in the study was
voluntary, and all participants had the option of withdrawing from the study at any time without
a penalty. Adolescents’ verbal agreement and parents’ informed written consent (forms were
disseminated by class teachers three weeks before the baseline assessment) were obtained before
conducting the study. After completion of the survey, the questionnaires were placed in an
envelope and sealed in front of the class. Students were assured that their individual information
61
would not be seen by parents or school administrators (Morgenstern et al., 2013). Data analyses
involved two assessment waves that took place approximately 12 months apart: baseline (Fall-
Spring of the school year 2018-2019), and 12-month follow-up (Fall-Spring of the school year
2019-2020). Study implementation was approved by the ministries of cultural affairs of the six
involved states. Ethical approval was obtained from the Ethical Committee of the German
Psychological Society (Ref. No. RH 042015_1).
Measures
The survey included the following items: (1) ED use behavior, (2) ADHD symptoms
frequency, (3) substance use behavior, (4) sociodemographic, (5) environmental, and (6)
intrapersonal (within-person/ individual characteristics) factors.
Energy Drink Use and ADHD symptoms. ED use was assessed at each time point. To
examine the lifetime ED use status, the following question was asked: “Have you ever tried
energy drinks (e.g., Red Bull, Monster) in your life?”. Response categories were “yes” or “no”.
Current ED consumption was assessed by asking, “How often do you currently drink energy
drinks?”. The response categories included “Not at all”, “Less than once a month”, “At least
once a month, but not every week”, “At least once a week, but not every day”, and “Almost
every day”. Answers other than “Not at all” and “Less than once a month” were coded as past
30-day use of EDs.
Self-reported ADHD symptoms were assessed using the 5-item hyperactivity-inattention
subscale (Goodman, 2001; Goodman et al., 2003; Muris et al., 2003) of the strengths and
difficulties questionnaire (SDQ). SDQ is a brief, 25-item survey of behavioral and emotional
difficulties, which can be utilized to assess mental health problems in children and adolescents
aged 4–17 years (Goodman & Goodman, 2009; Goodman et al., 2003). Numerous U.S. and
62
European studies confirmed the validity and reliability of the SDQ scale while also confirmed
that it is a valid tool to detect cases with ADHD among children and adolescents (Algorta et al.,
2016; Cuffe et al., 2005; Goodman et al., 2003; Hall et al., 2019). Respondents indicated the
applicability of each item statement or symptom (i.e., “I am restless, I cannot stay still for long”,
“I am constantly fidgeting or squirming”, “I am easily distracted, I find it difficult to
concentrate”) on a 3-point answer scale ranging from 0 = “not true”, 1 = “somewhat true”, and 2
= “certainly true”. The answers to these five items were further summed (Goodman et al., 2003)
in the self-reported ADHD symptoms frequency scale (Cronbach’s alpha=0.70). Further, the
scores from 0 to 6 were classified as “normal/borderline”, while the scores ranging 7 to 10 were
classified as “abnormal” (i.e., classified as having ADHD symptoms; Goodman, 2001; Goodman
et al., 2003; Muris et al., 2003). Self-reported ADHD symptoms measured at baseline was the
primary exposure variable, while lifetime use of ED at 12-month follow-up was the primary
outcome.
Substance use behavior. Lifetime substance (i.e., cigarettes, e-cigarettes, hookah, and
alcohol) use behavior was assessed with the following items: “In your lifetime, how many times
have you tried cigarettes /e-cigarettes or e-hookah/ hookah /alcohol?” There were five response
options for each question: “Never”, “Few puffs/ Just a little bit”, “1 to 19 times”, “20-100
times”, and “More than 100 times”. Answers other than “Never” (and “Just tried a little bit” for
alcohol use question) were coded as lifetime use of each corresponding product.
Covariates. Possible confounding influences were addressed. Baseline factors correlated
with ED use or ADHD symptoms in previous studies were included as covariates (Azagba et al.,
2014; De Alwis et al., 2014; Galimov et al., 2019; Gardner & Gerdes, 2015; Masengo et al.,
2020; Pan & Yeh, 2017; Terry-McElrath et al., 2014; Visram et al., 2016).
63
Sociodemographic. Age, gender, and socioeconomic status (SES) were assessed. The
latter was assessed (Goodman et al., 2001) with the item, “Please place an ‘X’ on the step that
best represents where you think your family stands on the ladder?” (on a ten-point scale
corresponding to a picture of a ladder, ranging from 0 [low income, the worst jobs, the lowest
education] to 10 [high income, the best jobs, the highest education]).
Environmental factors. Indicators of the proximal environment included the type of
school participants attended (i.e., gymnasium or other). The school type was coded as “1” if the
student attended a gymnasium (the most advanced type of secondary school in Germany,
strongly emphasizes academic learning) and “0” if the student attended another type of school.
Intrapersonal factors. Personality traits and psychological processes linked with ADHD
and ED use were assessed. Self-reported school performance (Achenbach, 1991; Sharif &
Sargent, 2006) was evaluated with the question: “How would you rate your school performance
compared to the classmates in your class?”. Responses corresponded to the following scores: 1-
“much worse”; 2 - “somewhat worse”; 3 - “about the same”; 4 - “somewhat better”; and 5 -
“much better”.
Risk-taking (Stephenson et al., 2003) was evaluated using a 2-item index, averaging
(Cronbach’s alpha = 0.9) responses to the following two questions: “How often do you do
dangerous things to have fun?” and “How often do you do exciting things, even if they are
dangerous?” (each on a 5-point scale ranging from “not at all” to “very often”).
Depression symptoms were assessed using the 5-item emotional problems subscale
(Goodman, 2001; Goodman et al., 2003; Muris et al., 2003) of the SDQ survey (“I worry a lot”,
“I am often unhappy, depressed or tearful”, “I am nervous in new situations”, “I easily lose
confidence”, “ I have many fears, I am easily scared”). The response options ranged from 0 =
64
“not true”, 1 = “somewhat true”, and 2 = “certainly true”. The answers to these five items were
further summed (Goodman et al., 2003) in the self-reported depression symptoms frequency
scale (Cronbach’s alpha=0.75)
Data Analysis
Participant accrual, sample size, and exclusions from the analytic sample are described
first. Then the correlates of study attrition are reported. Next, the prevalence, descriptive
statistics of the study variables for the total sample and stratified by lifetime ED are reported.
The correlates of a lifetime and past 30-day ED use measured at baseline among the full
sample (n=5,478) were assessed in two separate multilevel logistic regression models (ICCs:
0.12-0.16). Two separate multilevel models included baseline lifetime and past 30-day ED use as
dependent variables and the following independent variables: age, gender, type of school, self-
reported SES, school performance, risk-taking, self-reported ADHD and depression symptoms,
and lifetime use of cigarettes, e-cigarettes, and alcohol, accounting for the nesting of students
(Level 1) within eighty-three schools (Level 2). Further, the multilevel logistic regression model
was used (ICC=0.07) to test whether the baseline ADHD symptoms (coded as 0 -
“normal/borderline” symptoms, 1 - “abnormal/ADHD” symptoms) predicted ED use initiation in
12 months, accounting for the nesting of students (Level 1) within eighty-three schools (Level 2).
This analysis was conducted among baseline ED non-users (2,877). In a supplementary analysis,
this model was replicated among a subgroup of children aged 9-10 years. Finally, to test the
reversed temporal ordering, an additional multilevel model (ICC=0.04) was conducted. This
model used the baseline ED use variable as a predictor and ADHD symptoms measured at 12-
month follow-up as an outcome while controlling for baseline ADHD symptoms. All models
were fitted with and without adjustment for all covariates.
65
Maximum likelihood estimation was used to account for non-normal distributions and
missing data. Continuous variables were standardized (mean = 0, SD = 1) to facilitate
interpretation. All statistical analyses were conducted using Stata software (version 15.1; Stata
Corp, College Station, Texas, USA). Odds ratios (ORs) and Beta coefficients (βs) with 95% CIs
were reported with statistical significance set at P < .05 (2-tailed). Benjamini-Hochberg multiple
testing corrections were applied to control the false-discovery rate at .05. Additional sensitivity
analyses are summarized below in the Results section.
RESULTS
Study Sample
Participant accrual, sample size, and exclusions from the analytic sample are depicted in
Figure 6. Among 16,780 eligible participants, 14,242 (85.0%) provided parental consent and
were present on the day of the baseline survey. After excluding 406 (2.9%) students who did not
complete data on baseline key variables (i.e., baseline ED and substance use) and those students
who graduated from school after the 10th grade 4,265 (30.1%) of the 13,836 participants
administered the full-length baseline, 5,619 (58.7%) completed both assessments. After
excluding 141 (2.5%) adolescents who did not complete data on ADHD symptoms analytic
sample was 5,478.
66
Figure 6. The flow of participants in Study 2.
67
Attrition analysis
Baseline participants with and without follow-up data did not differ by baseline ED use
or any other covariate except for age and school type. That is, students without follow-up data
were older (p<0.001) and studied in schools other than gymnasia (p<0.001).
Descriptive Analyses
Among the 5,478 participants included in the sample, half (49.9%) were boys. The age
ranged between 9-18 years with a mean of 12.4 (SD=1.5) years. Roughly half of the students
(56.6%) attended gymnasia (high academic track schools). Of the total participants in the sample,
47.5% were lifetime ED users, and 13.9% were current ED users. Bivariate analysis
demonstrated that lifetime (4.6 [1.5] vs. 4.3 [1.4], p<0.001) as well as past 30-day ED users (4.6
[1.6] vs. 4.4 [1.5], p=0.001) reported more frequent ADHD symptoms at baseline compared to
ED non-users. Other baseline sociodemographic, environmental, and intrapersonal
characteristics for the total sample size and stratified by ED use status are reported in Table 8.
68
Table 8. Participant baseline characteristics for the total sample and by ED use.
Study variables
a
Total
sample
(n = 5478)
Lifetime ED use Past 30-day ED use
Users
(n=2601)
Non-users
(n=2877)
Users
(n=760)
Non-users
(n=4715)
Total 100 47.5 52.5 13.9 86,1
Sociodemographic factors
Gender
- Male 49.9 56.6 43.8 60.8 48.1
Age, mean (SD) 12.4 (1.5) 13.0 (1.4) 11.9 (1.4) 13.4 (1.3) 12.3 (1.5)
Self-reported SES, mean (SD) 6.8 (1.5) 6.7 (1.5) 6.9 (1.4) 6.7 (1.6) 6.9 (1.4)
Environmental Factors
School type
- Gymnasia 56.6 44.5 67.6 31.6 60.7
- Other 43.4 55.5 32.4 68.4 39.3
Intrapersonal Factors
Risk taking scale, mean (SD) 2.2 (1.0) 2.6 (1.1) 1.9 (0.9) 3.0 (1.1) 2.1 (1.0)
School performance, mean
(SD)
3.2 (0.8) 3.1 (0.8) 3.3 (0.8) 2.9 (0.9) 3.2 (0.8)
Depression symptoms scale,
mean (SD)
2.7 (2.3) 3.0 (2.4) 2.4 (2.1) 3.2 (2.5) 2.6 (2.2)
Substance use and ADHD
Baseline self-reported ADHD
symptoms scale, mean (SD)
4.5 (1.5) 4.6 (1.5) 4.3 (1.4) 4.6 (1.6) 4.4 (1.5)
Baseline lifetime use
- Cigarettes 13.7 26.7 1.9 47.5 8.2
- E-cigarettes 12.2 25.0 0.7 47.1 6.6
- Hookah 12.1 24.2 1.1 44.1 6.9
- Alcohol 23.5 41.8 7.0 60.0 17.6
Abbreviations: SES, socioeconomic status; SD, standard deviation
All pairwise comparisons between ED users and non-users were statistically significant at p<0.001.
a
Data are expressed as percent (%) unless otherwise indicated
69
Cross-sectional Analyses
The multilevel analysis revealed that after adjusting for other variables, baseline self-
reported ADHD symptoms frequency (Adjusted Odds Ratio [AOR]=1.09 [95% CI, 1.01-1.17]),
lifetime use of cigarettes (2.26 [95% CI, 1.59-3.21]), e-cigarettes (4.69 [95% CI, 2.81-7.80]),
hookah (4.30 [95% CI, 2.82-6.56]) and alcohol (2.89 [95% CI, 2.35-3.56]), greater risk-taking
(1.70 [95% CI, 1.56-1.85]) and depression scores (1.14 [95% CI, 1.05-1.23]), male gender (2.04
[95% CI, 1.75-2.38]), and older age (2.27 [95% CI, 2.05-2.52]) were associated with higher odds
of being a lifetime ED user. Meanwhile, attending a gymnasium-type school (0.45 [95% CI,
0.35-0.58]) and better school performance (0.81 [95% CI, 0.75-0.87]) were associated with lower
odds of being a lifetime ED user (see Table 9). The results of the multilevel model examining the
association between past 30-day ED use and the same set of independent variables showed
comparable results, except for baseline ADHD symptoms, which were no longer significant.
Prospective Analyses
There were 2,877 (52.5%) students who never tried EDs at baseline; in the subsequent 12
months, 761 (26.5%) of them had initiated (ever tried) EDs. The mean age of those initiated EDs
was 11.1 (SD=1.3) years. An unadjusted multilevel logistic regression model demonstrated that
in a sample of baseline ED non-users, having ADHD symptoms were not associated with ED use
initiation (OR=1.34 [95% CI, 0.98-1.84]) in a subsequent 12-month period (see Table 10).
Similarly, the covariate-adjusted model also failed a find a significant association between
baseline self-reported ADHD symptoms and ED use initiation (AOR=1.23 [95% CI, 0.89-1.71])
in 12 months. Nonetheless, parameter estimates for covariates in the adjusted models showed
that male gender, older age, baseline alcohol use, and higher risk-taking were positively
70
associated, while studying in the gymnasium was negatively associated with ED use initiation
(p<0.05, see Table 10).
Table 9. Multilevel models examining the cross-sectional associations between study variables
and lifetime/past 30-day ED use.
Independent variables Lifetime ED use Past 30-day ED use
AOR (95% CI) AOR (95% CI)
Baseline ADHD symptoms 1.09 (1.01-1.17)
*
1.03 (0.94-1.14)
Gender (female=ref) 2.04 (1.75-2.38)
*
1.95 (1.56-2.42)
*
School type (other=ref) 0.45 (0.35-0.58)
*
0.38 (0.28-0.51)
*
Age
a
2.27 (2.05-2.52)
*
1.56 (1.36-1.78)
*
Self-report SES
a
0.98 (0.91-1.06) 1.01 (0.92-1.11)
School performance
a
0.81 (0.75-0.87)
*
0.80 (0.72-0.88)
*
Risk-taking
a
1.70 (1.56-1.85)
*
1.66 (1.50-1.84)
*
Depression symptoms
a
1.14 (1.05-1.23)
*
1.13 (1.02-1.26)
*
Lifetime use (never use=ref)
- Cigarettes 2.26 (1.59-3.21)
*
1.67 (1.28-2.17)
*
- E-cigarettes 4.69 (2.81-7.80)
*
1.97 (1.48-2.61)
*
- Hookah 4.30 (2.82-6.56)
*
1.91 (1.46-2.50)
*
- Alcohol 2.89 (2.35-3.56)
*
2.18 (1.73-2.75)
*
Abbreviations: SES, socioeconomic status; SD, standard deviation AOR=Adjusted Odds Ratio; CI=Confidence
Interval; ref=reference
* Statistically significant after Benjamini-Hochberg corrections for multiple testing to control the false-discovery
rate at .05 (based on 2-tailed corrected P-value).
a
Rescaled (mean = 0, SD = 1) such that the ORs indicate the change in odds in the outcome associated with an
increase in 1 SD unit on the covariate continuous covariate scale
71
Table 10. Multilevel model examining the temporal associations between ED use and ADHD
symptoms.
Independent variables ED use at 12-month follow-up
OR (95% CI)
Unadjusted Models
Baseline ADHD symptoms
(normal/borderline=ref)
1.34 (0.98, 1.84)
Adjusted Models
Baseline ADHD symptoms
(normal/borderline=ref)
1.23 (0.89-1.71)
Gender (female=ref) 1.29 (1.06-1.56)
*
School type (other=ref) 0.53 (0.39-0.72)
*
Age
a
1.35 (1.19-1.53)
*
Self-report SES
a
1.04 (0.94-1.15)
School performance
a
0.94 (0.85-1.03)
Risk-taking
a
1.32 (1.19-1.48)
*
Depression symptoms
a
1.03 (0.93-1.15)
Lifetime use (never use=ref)
- Cigarettes 1.97 (1.04-3.70)
- E-cigarettes 0.35 (0.11-1.12)
- Hookah 2.03 (0.87-4.72)
- Alcohol 1.66 (1.20-2.32)
*
Abbreviations: OR= Odds Ratio; CI=Confidence Interval; ref=reference
* Statistically significant after Benjamini-Hochberg corrections for multiple testing to control the false-discovery
rate at .05 (based on 2-tailed corrected P-value).
a
Rescaled (mean = 0, SD = 1) such that the ORs indicate the change in odds in the outcome associated with an
increase in 1 SD unit on the continuous covariate scale
72
Supplementary Analyses
In a supplementary analysis, mentioned above procedures were replicated on a subset of
children (n=594) aged 9-10 years. There were 489 (82.3%) children aged 9-10 years who never
tried EDs at baseline; in the subsequent 12 months, 74 (15.1%) of them had initiated ED use. In
the unadjusted multilevel logistic regression model. An unadjusted multilevel logistic regression
model demonstrated that in a sample of baseline ED non-users (aged 9-10 years), those who had
ADHD symptoms were significantly associated with ED use initiation (4.30 [95% CI, 1.93-
9.55]) in subsequent 12-month period compared to those without ADHD symptoms. This
association also remained significant after adjustment for covariates (3.22 [95% CI, 1.43-7.22]),
as well as after applying the Benjamini-Hochberg correction for multiple comparisons (see Table
11).
An additional supplementary multilevel model was conducted to investigate the reversed
temporal ordering between ED use and ADHD symptoms (n=5478). The results of the
unadjusted multilevel logistic regression model demonstrated that after controlling for baseline
ADHD symptoms, baseline lifetime ED users (0.41 [95% CI, 0.29-0.52]) and current ED users
(0.79 [95% CI, 0.64-0.95]) had reported more frequent ADHD symptoms at 12-months follow-
up compared to ED non-users (see Table 12). This association remained significant after
adjustment for covariates (0.21 [95% CI, 0.08-0.33] for lifetime ED use versus non-use and 0.31
[95% CI, 0.12-0.50] for current ED use versus non-use), as well as after applying the Benjamini-
Hochberg correction for multiple comparisons.
In parallel (sensitivity) analyses missing data were handled using multivariate
imputations by chained equations (MICE) method for missing at random assumptions and the
available covariate data, 25 imputed data sets were created. The pooling of the regression
73
estimates followed Rubin’s rule (White et al., 2011). All significant differences in these analyses
remained and were in the same direction.
Table 11. Multilevel model examining the temporal associations between ED use and ADHD
symptoms among 9–10-years-old children.
Independent variables ED use at 12-month follow-up
OR (95% CI)
Unadjusted Models
Baseline ADHD symptoms
(normal/borderline=ref)
4.30 (1.93-9.55)
*
Adjusted Models
Baseline ADHD symptoms
(normal/borderline=ref)
3.22 (1.43-7.22)
*
Gender (female=ref) 1.39 (0.79-2.46)
School type (other=ref) 0.34 (0.18-0.63)
*
Age
a
-
Self-report SES
a
1.07 (0.81-1.42)
School performance
a
0.97 (0.68-1.37)
Risk-taking
a
1.58 (1.17-2.13)
*
Depression symptoms
a
1.20 (0.88-1.62)
Lifetime use (never use=ref)
- Cigarettes -
- E-cigarettes -
- Hookah -
- Alcohol -
Abbreviations: OR= Odds Ratio; CI=Confidence Interval; ref=reference
* Statistically significant after Benjamini-Hochberg corrections for multiple testing to control the false-discovery
rate at .05 (based on 2-tailed corrected P-value).
a
Rescaled (mean = 0, SD = 1) such that the ORs indicate the change in odds in the outcome associated with an
increase in 1 SD unit on the covariate continuous covariate scale
74
Table 12. Multilevel models examining the reversed temporal ordering between ED use and
ADHD symptoms
Independent variables ADHD symptoms at 12-month follow-up (n=5478)
β (95% CI)
Unadjusted Models
Baseline ADHD symptoms scale
a
0.61 (0.56, 0.66)
*
Baseline ED use (never use=ref)
- Lifetime use
b
0.41 (0.29, 0.52)
*
- Past 30-day use 0.79 (0.64, 0.95)
*
Adjusted Models
Baseline ADHD symptoms scale
a
0.51 (0.46, 0.56)
*
Baseline ED use (never use=ref)
- Ever use 0.21 (0.08, 0.33)
*
- Past 30-day use 0.31 (0.12, 0.50)
*
Gender (female=ref) 0.04 (-0.07, 0.14)
School type (other=ref) -0.17 (-0.34, -0.01)
Age
a
-0.22 (-0.29, -0.15)
*
Self-report SES
a
-0.08 (-0.14, -0.03)
*
School performance
a
-0.40 (-0.46, -0.35)
*
Risk-taking
a
0.32 (0.26, 0.38)
*
Depression symptoms
a
0.28 (0.23, 0.34)
*
Lifetime use (never use=ref)
- Cigarettes 0.18 (-0.01, 0.37)
- E-cigarettes -0.07 (-0.28, 0.15)
- Hookah 0.07 (-0.13, 0.27)
- Alcohol 0.05 (-0.09, 0.20)
Abbreviations: SES, socioeconomic status; β = Betta coefficient; CI=Confidence Interval; ref=reference
* Statistically significant after Benjamini-Hochberg corrections for multiple testing to control the false-discovery
rate at .05 (based on 2-tailed corrected P-value).
a
Rescaled (mean = 0, SD = 1) such that one unit change β is associated with an increase in 1 SD unit on the
continuous covariate scale
b
This category includes lifetime ED users but excludes past 30-day ED users
75
DISCUSSION
This study examined the temporal associations between ED use behavior and self-
reported ADHD symptoms by analyzing data from a large German school-based adolescent
sample. Lifetime ED use was reported by 47.5% of participants, while 13.9% reported current
ED use. During a 12-month period, 26.5% of the participants who never tried EDs at baseline,
initiated ED use (mean age 11.1 years). Consistent with the previous longitudinal study
(Galimov et al., 2019), ED use initiation was associated with male gender, older age, higher risk-
taking, and inversely associated with attending high academic track schools (gymnasia). Baseline
self-reported ADHD symptoms failed to predict subsequent ED use initiation in the next 12
months; however, this association was significant among the subgroup of children aged 9-10
years. Additionally, after controlling for baseline ADHD symptoms, ED use at baseline predicted
more frequent ADHD symptoms at 12-months follow-up.
Contrary to the first study hypothesis, baseline self-reported ADHD symptoms were not
predictive of subsequent ED use initiation; however, a subgroup analysis conducted among 9-10
years old children detected a significant association between these factors. Previous studies
reported that the average age of initial ED consumption ranges between 10-15 years (Costa et al.,
2016; Reid et al., 2015; Trapp et al., 2020). Moreover, growing evidence suggests that ED
marketing, one of the leading factors associated with ED initiation (Galimov et al., 2019),
became more youth-oriented in recent years (Hammond & Reid, 2018; Harris, 2013; Harris &
Munsell, 2015; Stacey et al., 2017). Thus, it is possible that mentioned above and other factors
(i.e., peer use) distorted the hypothesized link between ADHD symptoms and ED use initiation
and did not allow to detect a significant effect in the full analytic sample. Additionally, schools
included in our sample excluded schools offering “special education”; thus, it is possible that
76
adolescents with the most prominent ADHD symptoms were not included in our sample. Further,
restricting the study sample to children aged 9-10 years minimized the confounding influence of
these factors, which resulted in a significant association between baseline self-reported ADHD
symptoms and subsequent ED use initiation. This finding may be explained by several
mechanisms. First, there may be a common factor linking ADHD symptoms and ED use. For
instance, ADHD adolescents may experience a low level of parental involvement (Rogers et al.,
2009) or get inadequate sleep and have higher risk-taking traits compared to adolescents without
ADHD symptoms (Konofal et al., 2010; Pollak et al., 2019), which might predispose them to try
EDs. It is also possible that adolescents with ADHD symptoms are more susceptible to ED
marketing than non-ADHD adolescents, making them more prone to try EDs.
Alternatively, one can speculate that the link between ADHD symptoms and ED
initiation could be explained by the self-medication hypothesis (Krause et al., 2002; Lawrence et
al., 2002; Levin et al., 2006). Previous studies have demonstrated that dopamine reinforcement
dysfunction and abnormalities in the structure of critical brain regions related to dopamine are
the primary causes of ADHD (Swanson et al., 2007; Volkow et al., 2009). Further, EDs contain
caffeine, guarana, ginseng, and kola nut, which are known for their strong stimulating effect
(Reissig et al., 2009; Vercammen et al., 2019). Thus, given the increasing popularity of EDs
among youth, lack of minimum age of sale, as well as their wide availability, children and
adolescents (especially those who don’t have access to other substances, such as cigarettes, e-
cigarettes, and alcohol) with ADHD symptoms may choose to self-medicate with EDs in order to
cope with their symptoms (Krause et al., 2002; Lawrence et al., 2002; Levin et al., 2006).
In line with the second study hypothesis, ED use at baseline (controlling for baseline
ADHD symptoms) predicted more frequent ADHD symptoms at 12-months follow-up. This
77
association could be explained by common factors, such as inadequate sleep, predisposition to
sensation seeking, and low level of parental monitoring (Konofal et al., 2010; Pollak et al., 2019;
Rogers et al., 2009). To address the possible influence of common risk factors, the multilevel
model was adjusted for sociodemographic, environmental, and intrapersonal characteristics that
potentially could overlap with ED use and ADHD. After adjustment for these factors, the OR
estimates associated with ED use reduced but remained significant. Further, the link between ED
use and ADHD symptoms could be explained by neurobiological mechanisms. For instance, it is
possible that, unlike other stimulant use, ED use behavior may cause dysregulation of
mesolimbic function, which in turn may lead to increased signs and symptoms of ADHD. As
was mentioned earlier, dopamine reinforcement dysfunction and abnormalities in the structure of
critical brain regions related to dopamine are one of the primary causes of ADHD (Swanson et
al., 2007; Volkow et al., 2009). Numerous studies suggest that caffeine contained in EDs has a
stimulating effect on mesolimbic reward pathways of the brain and can increase the release of
dopamine (Nall et al., 2016; Nehlig et al., 1992; Solinas et al., 2002). Thus, one can argue that
the dysregulation of mesolimbic function caused by ED use in children and adolescents with
ADHD symptoms may further lead to exacerbation of signs and symptoms of ADHD (Tripp &
Wickens, 2009).
Limitations
This study is unique in its longitudinal examination of ED use and ADHD symptoms
among a large nationally representative sample of German children and adolescents. However,
several limitations should be taken into account. Loss to follow-up could have affected
generalizability as older students and those who studied in schools other than gymnasia were
more likely to drop out from the study; thus, the study results may have limited generalizability
78
to high-risk adolescents. Some of the predictors were single-item measures; thus, future studies
should include multiple items to assess the construct of interest to reduce measurement error and
increase reliability. While this study focused on ED use, overall and exact amounts of caffeine
consumption were not measured. Given that EDs may not be the primary caffeine source for
children and adolescents (Verster & Koenig, 2018), further research should account for other
sources of caffeine to rule out possible confounding. Given the nature of the data (self-reported),
recall and social desirability, and cognitive biases may have affected the results in both
adolescent samples. For instance, the ADHD symptoms measure was a self-reported item, hence
it is possible that individuals with a diagnosis of ADHD differed in the severity of illness from
those in clinical studies. The first study hypothesis was only confirmed among the subgroup of
children aged 9-10 years, hence future studies should include larger and more diverse study
samples to achieve sufficient power to detect the significant effect; one should also control for
other important confounders. For instance, some other important covariates, such as marketing
exposure, peer ED use, and family environment (which could have distorted the results) were not
assessed in this study and should be included in future work.
Conclusions
The results of this study indicate that having ADHD symptoms is associated with ED use
initiation in 12 months. Additional studies involving larger and more diverse samples are needed
to better understand the mechanism behind this association. Given that ED use is associated with
subsequent substance use initiation (see Chapter 3), delivery of interventions aimed at teaching
“healthy” coping strategies may help to reduce the risk for later substance use initiation for
children and adolescents with ADHD diagnoses who consume EDs to cope with their symptoms.
Further, it was demonstrated that ED consumption is a risk marker for increasing ADHD
79
symptoms over time. Thus, regular screening for ED use is warranted for youth with prevalent
self-reported ADHD symptoms. Additionally, early screening and diagnosis for ADHD disorder,
as well as timely medication treatment, may help prevent the self-medication of EDs and other
substances among children and adolescents with self-reported is ADHD symptoms.
80
Chapter 3: ADHD Symptoms and Substance Use Among German Adolescents: Energy
Drink Consumption as A Risk Factor.
INTRODUCTION
Attention-deficit/hyperactivity disorder (ADHD) is a prevalent psychiatric condition
characterized by persistent symptoms of impulsivity, hyperactivity, and difficulty sustaining
attention (American Psychiatric Association, 2013). An estimated overall pooled prevalence of
ADHD among youth, reported in a recent systematic review that evaluated results of 175 studies,
was 7.2 percent (Thomas et al., 2015). However, evidence suggests that this rate is possibly
increasing in certain populations (Nyarko et al., 2017; Visser et al., 2010). The etiology of
ADHD involves a combination of genetic (disorder tends to run in families),
neurodevelopmental, and environmental factors (Thapar et al., 2013). Additionally, evidence
indicates that increased exposure to the latter modifiable risk factors may be associated with
growing ADHD prevalence in some populations (Larson et al., 2014).
ADHD is a major clinical and public health problem because of its associated morbidity
and disability in children, adolescents, and adults (Goldman et al., 1998). A few studies have
indicated that adolescents with a formal ADHD diagnosis, as well as those with self-reported
ADHD symptoms, experience educational problems, report poor academic performance and peer
relationship difficulties (Barry et al., 2002; Diamantopoulou et al., 2005; Gardner & Gerdes,
2015; Rapport et al., 1999; Strine et al., 2006). Moreover, a wide range of comorbid psychiatric
disorders are associated with ADHD (in clinical and community samples), including oppositional
defiant disorder, conduct disorder, anxiety, and depression (Biederman, 2005; Jensen et al.,
2001; Pan & Yeh, 2017; Yen et al., 2007). Serval longitudinal studies have demonstrated that
adolescents with ADHD are more likely to experience an early onset of substance use disorders
81
than non-ADHD individuals (Biederman et al., 1997; Katusic et al., 2005; Wilens et al., 2011).
While most of these studies have been conducted among clinical adolescent samples, a growing
body of evidence suggests that self-reported ADHD symptoms are associated with substance use
behavior in general population samples. For instance, it was shown that self-reported ADHD
symptoms among adolescents with no formal ADHD diagnosis are associated with substance
(alcohol, nicotine, marijuana, cocaine) use behaviors (De Alwis et al., 2014; Kessler et al., 2012).
Neurobiological and neuropsychological studies attempted to explain the relationship
between substance use and ADHD. Few studies suggested that dopamine reinforcement
dysfunction and abnormalities in the structure of critical brain regions related to dopamine are
primary causes of ADHD (Swanson et al., 2007; Volkow et al., 2009). On the other hand, a few
other studies have shown that smoking, alcohol, and psychostimulant medications use can cause
a stimulating effect on dopamine release and utilization (Brody et al., 2009; Krause et al., 2002;
Ostroumov et al., 2015; Potter & Newhouse, 2008). This is not surprising as it is believed that
dysfunctions in mesocortical brain networks and the mesolimbic dopaminergic system are the
primary causes of ADHD (Biederman, 2005; Swanson et al., 2007). Hence, consistent with the
self-medication hypothesis, it is possible that adolescents with ADHD may self-medicate with
nicotine, alcohol, and other stimulating substances to manage their symptoms (Krause et al.,
2002; Lawrence et al., 2002; Levin et al., 2006).
Caffeine is another substance that has a stimulating effect on mesolimbic reward
pathways of the brain and can increase the release of dopamine (Nall et al., 2016; Nehlig et al.,
1992; Solinas et al., 2002). Caffeine consumption is a socially accepted behavior in many
countries and is very prevalent worldwide (Meredith et al., 2013). One of the beverages that
contain high levels of caffeine along with other stimulating ingredients (such as guarana and
82
taurine) and that have become extremely popular among youth are energy drinks (EDs) (Emond
et al., 2015; Facts, 2013; Reissig et al., 2009). Given the strong stimulating effects of ED
ingredients coupled with the wide availability of these beverages and no minimum age of sale
restrictions, in line with the self-medication hypothesis (Krause et al., 2002; Lawrence et al.,
2002; Levin et al., 2006), it is possible that adolescents with ADHD may self-medicate with EDs
to cope with their symptoms. In this line, one cross-sectional study and Study 2 of this
dissertation observed an association between ED consumption and self-reported ADHD
symptoms (Schwartz et al., 2015).
Previous studies involving adolescents have shown that ED consumption is associated
with poor academic achievement, school stress, depression, and sensation seeking (Azagba et al.,
2014; Galimov et al., 2019; Masengo et al., 2020; Park et al., 2016; Poulos & Pasch, 2015). ED
use has also been linked to risky behaviors, such as substance use (Choi et al., 2016; Galimov et
al., 2019; Miyake & Marmorstein, 2015). Moreover, as demonstrated in Chapter 2 of this
dissertation, baseline ED use among substance non-users is associated with tobacco, e-cigarette,
alcohol, and marijuana use initiation in 12 months. Thus, given the prospective association
between ADHD symptoms (clinical and subclinical) and substance use among adolescents
(Biederman et al., 1997; De Alwis et al., 2014; Katusic et al., 2005; Kessler et al., 2012; Wilens
et al., 2011), it is also plausible that ED use mediates the association between ADHD symptoms
and substance use initiation.
This study extends previous work and fills a significant research gap by examining the
mediating role of ED consumption in the association between self-reported ADHD symptoms
among baseline substance non-users and smoking (cigarettes, e-cigarettes, hookah), alcohol, and
marijuana use initiation. It is hypothesized that the association between baseline ADHD
83
symptoms and tobacco, e-cigarette, alcohol, and marijuana use initiation in 12 months is
mediated by baseline ED use frequency. The heuristic model of this study is summarized in
Figure 7.
METHODS
Participants and procedures
Data were collected as part of an ongoing longitudinal cohort survey of substance use and
risk factors, conducted among adolescents from six Federal states of Germany: Baden-
Württemberg, Mecklenburg-West-Pomerania, North-Rhine-Westphalia, Rhineland-Palatinate,
Saxony, and Schleswig-Holstein. These states were randomly selected from one of six Nielsen
regions, which cluster areas with similar purchasing power and consumer behavior (GmbH,
2016). A total of 627 schools representing all types of schools except for schools for students
with special needs were identified in randomly selected sub-regions within each state and invited
to participate in the study (Hansen et al., 2018). Eighty-three of these schools signed in and
returned a registration form indicating their intention to participate in the study.
84
Figure 7. Study 3 heuristic model.
Data were collected through self-completed anonymous questionnaires. Assessments
were carried out during one school hour (45 min) by trained research staff or school personnel.
Adolescents’ verbal assent and parents’ written informed consent were obtained prior to
conducting the study. Data analyses involved two assessment waves that took place
approximately 12 months apart: baseline (Fall-Winter of the school year 2016-2017) and 12-
month follow-up. To link the baseline and follow-up questionnaires, students were asked to
generate an anonymous seven-digit individual code, a procedure that had been tested in previous
85
studies, slightly modified for this study (Galanti et al., 2007; Hanewinkel et al., 2011;
Morgenstern et al., 2013).
Participation in the study was voluntary, and all participants had the option of
withdrawing from the study at any time without a penalty. After completing the survey, the
questionnaires were placed in an envelope and sealed in front of the class (Morgenstern et al.,
2013). Students were assured that parents or school administrators would not see their individual
information. Study implementation was approved by the ministries of cultural affairs of the six
involved states. Ethical approval was obtained from the Ethical Committee of the German
Psychological Society (Ref. No. RH 042015_1).
Measures
ADHD Symptoms. Self-reported ADHD symptoms were assessed using the 5-item
hyperactivity-inattention subscale (Goodman, 2001; Goodman et al., 2003; Muris et al., 2003) of
the strengths and difficulties questionnaire (SDQ). SDQ is a brief, 25-item survey of behavioral
and emotional difficulties, which can be utilized to assess mental health problems in children and
adolescents aged 4–17 years (Goodman & Goodman, 2009; Goodman et al., 2003). Numerous
U.S. and European studies confirmed the validity and reliability of the SDQ scale while also
confirmed the SDQ as a valuable tool to detect cases with ADHD among children and
adolescents (Algorta et al., 2016; Cuffe et al., 2005; Goodman et al., 2003; Hall et al., 2019).
Respondents indicated the applicability of each item statement or symptom (i.e., “I am restless, I
cannot stay still for long,” “I am constantly fidgeting or squirming,” “I am easily distracted, I
find it difficult to concentrate”) on a 3-point answer scale ranging from 0 = “not true,” 1 =
“somewhat true,” and 2 = “certainly true”. The answers to these five items were further summed
(Goodman et al., 2003) in the self-reported ADHD symptoms frequency scale (Cronbach’s
86
alpha=0.7). Self-reported ADHD symptoms index measured at baseline was the primary
exposure variable, while lifetime use of ED at 12-month follow-up was the primary outcome.
ED, Tobacco Use, and Alcohol Use Behavior. At each assessment wave, ED use
frequency was measured using the following question: “How often do you currently drink energy
drinks?” (“not at all,” “less than once a month,” “at least once a month, but not every week,” “at
least once a week, but not every day,” and “almost every day”).
Tobacco and alcohol use behaviors were also assessed at both waves with the questions:
“Have you ever tried cigarettes/ e-cigarettes/ hookah/ alcohol in your life?”. The response
categories included “never,” “a few puffs,” “1 to 19 times”, “20–100 times”, and “more than 100
times” for tobacco use items, and “never,” “just tried a little bit,” “1 time”, “2 to 5 times”, and
“more than five times” for alcohol use item. All answers other than “never” were coded as
lifetime cigarette/ e-cigarette/ hookah use, while response answers other than “never” and “just
tried a little bit” were coded as lifetime alcohol use.
Baseline ED use frequency (recoded as 0= not at all, 1= less than once a month, 2=at least
once a month, but not every week, 3= at least once a week or more frequently) and ADHD
symptoms were the primary exposure variables. The primary outcomes measured at one-year
follow-up were lifetime use of (1) cigarettes (yes or no); (2) e-cigarettes (yes or no); (3) hookah
(yes or no); and (4) alcohol (yes or no).
Covariates. Possible confounding influences were addressed. Baseline factors correlated
with ED use and substance use (i.e., cigarettes, e-cigarettes, hookah, and alcohol) or ADHD
symptoms in previous studies were included as covariates (Azagba et al., 2014; Covey & Tam,
1990; De Alwis et al., 2014; Fisher et al., 2007; Galimov et al., 2019; Gardner & Gerdes, 2015;
Masengo et al., 2020; Pan & Yeh, 2017; Terry-McElrath et al., 2014; Tyas & Pederson, 1998;
87
Visram et al., 2016). Sociodemographic, environmental, and intrapersonal (within-person/
individual characteristics) covariates were assessed.
Sociodemographic. Self-reported sociodemographic covariates were age, gender, and
SES measured (Goodman et al., 2001) with the question, “Please place an ‘X’ on the step that
best represents where you think your family stands on the ladder?”. Response options for this
item were on a ten-point scale corresponding to a picture of a ladder, ranging from 0 (low
income, the worst jobs, the lowest education) to 10 (high income, the best jobs, the highest
education).
Environmental Factors. School type and peer substance use were assessed as an
indicator of the proximal environment. The German school system has several types of
secondary schools (i.e., Hauptschule, Realschule, Oberschule, Gemeinschaftsschule,
Gymnasium) that differ with regard to the academic skills of their students and graduation level
(i.e., students typically graduate from school after 10
th
– 13
th
grade, depending on school type).
Gymnasia are the most advanced type of secondary school that strongly emphasizes academic
learning. The school type was coded as “1” if the student attended a gymnasium and “0” if
attended another type of school.
Peer substance use was assessed by responses to the question, “How many of your friends
use cigarettes/ e-cigarettes/ hookah/ alcohol?”. Response options for each question were on an
eleven-point scale ranging from “none” to “all”. Each response was dichotomized considering
answers of anything other than “none” as peer cigarette/ e-cigarette/ hookah/ alcohol use.
Further, these items were averaged (Cronbach’s alpha=0.83) in the peer substance use index.
Intrapersonal Factors. Personality traits and psychological processes linked with
experimentation, risky behavior, and substance use were assessed. Risk-taking (Stephenson et
88
al., 2003) was measured by the following two questions: “How often do you do dangerous things
to have fun?” and “How often do you do exciting things, even if they are dangerous?” Response
options for each statement were on a five-point scale ranging from “not at all” to “very often”.
Self-reported school performance (Achenbach, 1991; Sharif & Sargent, 2006) was
assessed by asking, “How would you rate your school performance compared to the classmates
in your class?” Response options included: “much worse,” “somewhat worse,” “about the same,”
“somewhat better,” and “much better”.
Depression symptoms were assessed using 5-item emotional problems subscale
(Goodman, 2001; Goodman et al., 2003; Muris et al., 2003) of SDQ assessment tool (“I worry a
lot,” “I am often unhappy, depressed or tearful,” “I am nervous in new situations,” “I easily lose
confidence,” “I have many fears, I am easily scared”). The response options ranged from 0 =
“not true,” 1 = “somewhat true,” and 2 = “certainly true”. The answers to these five items were
further summed (Goodman et al., 2003) in the self-reported depression symptoms frequency
scale (Cronbach’s alpha=0.80).
Data Analysis
Participant accrual and correlates of study attrition are reported. Then the prevalence of
ED use behavior and descriptive statistics are reported.
Multilevel structural equation modeling (SEM) was used to test the mediating effect of
ED use frequency on the relationship between self-reported ADHD symptoms and substance use
initiation. In this path model, baseline ADHD symptoms scale was used as exogenous variable,
ED use frequency as the mediator, and substance use initiation (latent variable) at 12-month
follow-up as the criterion, while accounting for potential clustering of the students (Level 1)
within eighty-three schools (Level 2). The model was conducted among baseline substance non-
89
users. The final model adjusted for all relevant covariates. The maximum likelihood method was
employed for parameter estimation. The significance of the indirect effect was tested using the
bootstrapping method. A 95% bias-corrected confidence interval was generated by bootstrapping
with 1000 re-samples. If the 95% confidence interval (CI) excluded zero, the indirect effect was
considered significant.
Data cleaning, descriptive analyses, and testing for moderating effects were conducted
using Stata software (version 15.1; Stata Corp, College Station, Texas, USA). SEM models were
conducted in Mplus Version 8.3. Continuous variables were standardized (mean = 0, SD = 1) to
facilitate interpretation. Beta coefficients (b) with 95% CIs were reported with statistical
significance set at P < .05 (2-tailed). Benjamini-Hochberg multiple testing corrections were
applied to control the false-discovery rate at .05.
RESULTS
Study Sample
Participant accrual, sample size, and exclusions from the analytic sample are depicted in
Figure 8. Among 16,780 eligible students, 14 242 (85.0%) provided parental consent and were
present on the day of the baseline survey. After excluding 406 (2.9%) participants who did not
complete data on baseline key variables (i.e., baseline ED and substance use) and those students
who graduated from school after the 10th grade 4,265 (30.1%) of the 13,836 participants (cross-
sectional analytic sample) administered the full-length baseline survey, 5 619 (58.7%) completed
both assessments. After excluding 168 (3.0%) adolescents who did not complete data on ADHD
symptoms follow-up substance use and1681 (29.9%) students who already used tobacco and
alcohol at baseline, the longitudinal analytic sample was 3770.
90
Attrition analysis
Baseline substance non-users with and without follow-up data did not differ by baseline
ED use or any other covariate except for age and school type. That is, students without follow-up
data were older (p<0.001) and studied in schools other than gymnasia (p<0.001).
Table 13. Participant characteristics.
Study variables
a
Total Sample
(n=3770)
Sociodemographic factors
Gender
- Male 49.4
- Female 50.6
Age, mean (SD) 12.0 (1.3)
Self-reported SES, mean (SD) 6.9 (1.4)
Environmental Factors
School type
- Gymnasia 62.4
- Other 37.6
Peer substance use
- Cigarettes 19.5
- E – cigarettes 13.8
- Hookah 16.7
- Alcohol 31.8
Intrapersonal Factors
Risk-taking scale, mean (SD) 2.0 (1.0)
Depression symptoms scale, mean (SD) 2.4 (2.1)
Self-reported school performance, mean (SD) 3.2 (0.8)
Substance use and ADHD
Baseline self-reported ADHD symptoms scale, mean (SD) 4.4 (1.5)
Lifetime substance use at follow-up
- Cigarettes 6.5
- E – cigarettes 6.1
- Hookah 5.5
- Alcohol 18.5
Abbreviations: SES, socioeconomic status; SD, standard deviation
a
Data are expressed as percent (%) unless otherwise indicated
91
Figure 8. The flow of participants in Study 3.
92
Descriptive Analyses
Among the 3,770 students included in the analytic sample, half (49.4%) were boys, while
the age ranged between 9-18 years with a mean of 12.0 (SD=1.3) years. About two-thirds of the
participants (62.4%) attended high academic track schools (gymnasia). Of the total participants
in the sample, 85.4% have never used EDs, 10.2% used them less than once a month, 2.8% used
EDs at least once a month, while 1.6% used EDs at least once a week or more frequently.
There were 3,770 students who never used substances (cigarettes, e-cigarettes, hookah,
alcohol) at baseline; in the subsequent 12 months, 6.5% of them had initiated cigarette use, 6.1%
e-cigarette use, 5.5% hookah use, and 18.5% alcohol use. Other baseline sociodemographic,
environmental, and intrapersonal characteristics are reported in Table 13.
Multilevel SEM
The multilevel mediational model (see Figure 9) demonstrated a good fit (RMSEA=.06;
CFI=.86; SRMR=.06). After controlling for sociodemographic, environmental, and intrapersonal
covariates, baseline self-reported ADHD symptoms were not directly associated with substance
use initiation at 12-month follow-up (β=.01, p=.236). Nonetheless, the multilevel SEM model
demonstrated that baseline ED use mediated the association between baseline self-reported
ADHD symptoms and substance use initiation at 12-month follow-up. In particular, baseline
self-reported ADHD symptoms frequency were positively associated with baseline ED use
behavior (β=.04, p<.001), which, in turn, was positively associated with substance use initiation
in subsequent 12 months (β=.06, p<.001).
93
Figure 9. Multilevel mediational model.
*The model is adjusted for sociodemographic, environmental, and intrapersonal covariates.
DISCUSSION
This study examined the mediating role of ED consumption in the association between
baseline self-reported ADHD symptoms and substance use initiation by analyzing data from a
large German school-based adolescent sample. Consistent with the study hypothesis, self-
reported ADHD symptoms appear to be associated with the risk of cigarette, e-cigarette, hookah,
and alcohol use initiation over a subsequent 12-month period through ED use mediational
pathway. Specifically, self-reported ADHD symptoms were positively associated with ED use
behavior, which was associated with substance use initiation.
The current findings are generally consistent with serval longitudinal studies suggesting
that adolescents with ADHD symptoms are more likely to experience an early onset of substance
use disorders than non-ADHD individuals (Biederman et al., 1997; Katusic et al., 2005; Wilens
et al., 2011). These results are also in line with population-based studies showing that self-
reported ADHD symptoms among adolescents with no formal ADHD diagnosis are associated
94
with substance (alcohol, nicotine, marijuana, cocaine) use behaviors (De Alwis et al., 2014;
Kessler et al., 2012). Additionally, the results of this study are consistent with the findings
observed in Chapter 3 of this dissertation that self-reported ADHD symptoms are associated with
ED use. Moreover, previous studies demonstrated that ED use is linked to impulsivity and
sensation-seeking, while a few studies proposed that ED use may be a marker for risk-taking
individuals in general (Azagba et al., 2014; Galimov et al., 2019; Miller et al., 2018). Thus, the
results of this study are also consistent with the previous mediation study showing that
impulsivity mediates the association between ADHD symptoms and substance use behaviors
(Egan et al., 2017).
Findings from this dissertation study are also in line with several theories that may
explain the observed association between ADHD symptoms, ED consumption, and substance use
behaviors. Dopamine reinforcement dysfunction and abnormalities in the structure of critical
brain regions related to dopamine are cited as one the primary causes of ADHD (Swanson et al.,
2007; Volkow et al., 2009). Hence, a few researchers proposed that adolescents with ADHD may
self-medicate with nicotine, alcohol, or EDs (see Chapter 3) to cope with their symptoms. It is an
example of the self-medication hypothesis (Krause et al., 2002; Lawrence et al., 2002; Levin et
al., 2006). Consistent with this hypothesis, the findings of this study have also demonstrated the
link between ADHD self-reported symptoms and ED use. Further, the observed association
between ED consumption and substance use initiation is in line with the gateway hypothesis,
which suggests that there are developmental stages of drug use among youth (see Chapter 1;
Kandel, 1975; Kandel et al., 1992; Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli,
2016; Vanyukov et al., 2012; Wagner & Anthony, 2002). Previous studies have also suggested
that the gateway hypothesis is particularly robust for children and adolescents with ADHD; for
95
instance, it was suggested that ADHD cigarette users compared to non-users are more likely to
initiate alcohol and other drugs as well as develop comorbid substance use disorders (Biederman
et al., 2006). Finally, observed results are consistent with cross-sensitization theory. According
to this theory, repeated consumption of a specific substance causes dysregulation in mesolimbic
structures of the brain and enhances the response to that substance, which may further produce a
hyper-responsive reaction to the other substances that act at the same neurobiological sites,
including ones that may have never been previously consumed (i.e., cross-sensitization; Hellberg
et al., 2019; Robinson et al., 1985; Sussman, 2017; Temple, 2009). Thus, one can speculate that
dysregulation in mesolimbic structures caused by ED use in children and adolescents with
ADHD symptoms (to manage their symptoms) may further lead to pathological motivation for
EDs, which in turn enhances the response to other substances that have never been previously
consumed.
The current study yields findings that may have important implications for prevention.
This study indicates that limiting ED use may contribute to a decreased risk for substance use
initiation among youth with ADHD symptoms. Hence, adopting policies that would restrict the
sale of EDs to minors may be helpful. For instance, EDs are not sold to adolescents under 15
years old in Sweden, while Denmark has banned their sales completely (Oddy & O'sullivan,
2009). Additionally, new policies banning the marketing of EDs and other substances to youth
under 18 years of age, including in traditional and social media, sponsorships, and other
activities with youth audiences, are warranted. It is also crucial for parents, school officials, and
healthcare providers to be aware of signs of excessive EDs consumption and limit their use by
adolescents (Harris & Munsell, 2015). Further, behavioral interventions aimed at preventing the
emergence of ED use behavior may reduce the risk for later substance use. For instance,
96
promotion of healthy alternative outlets for sensation-seeking (i.e., sports participation) and
teaching new “healthy” coping skills (mindfulness meditation) to manage ADHD symptoms may
lower the rates of substance use among children and adolescents with ADHD.
Limitations
Loss to follow-up could have affected generalizability as older students and those who
studied in low-middle academic track schools were more likely to drop out; thus, the results may
have limited generalizability to high-risk adolescents. Some of the predictors were single-item
measures, thus future studies should include multiple items to assess the construct of interest to
reduce measurement error and increase reliability. While this study focused on ED use, overall
and exact amounts of caffeine consumption were not measured. Given that EDs may not be the
primary caffeine source for children and adolescents (Verster & Koenig, 2018), further research
should account for other sources of caffeine to rule out possible confounding. Given the nature of
the data (self-reported), recall, social desirability, and cognitive biases may have affected the
results in both adolescent samples. For example, the ADHD symptoms variable was a self-
reported item, hence it is possible that individuals with a diagnosis of ADHD differed in the
severity of illness from those in clinical studies. Finally, other important covariates, such as
marketing exposure, peer ED use, and family environment, were not assessed in this study and
should be included in future work.
Conclusions
The results of this study indicate that one possible explanation why children and
adolescents with high levels of ADHD symptoms are at increased risk for substance use is ED
use behavior. The findings of this study suggest that reducing ED use may help curb the risk for
substance use initiation among youth with ADHD symptoms. ADHD youth are already at risk
97
for several detrimental outcomes as they mature into adulthood. Thus, this study's results may
help inform prevention and intervention programs that serve to decrease substance use behaviors
among these individuals.
98
CHAPTER 4: OVERALL DISCUSSION & CONCLUSION
Key findings from these studies contribute to a more comprehensive understating of ED
use behavior and its associated consequences, particularly among youth with self-reported
ADHD symptoms. This dissertation utilized data from two large, school-based adolescent
samples (U.S. and German) and attempted to integrate the problem behavior theory (Donovan et
al., 1991; Jessor, 1991; Jessor et al., 2003), cross-sensitization (Hellberg et al., 2019; Robinson et
al., 1985; Sussman, 2017; Temple, 2009), gateway hypothesis (Kandel, 1975; Kandel et al.,
1992; Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli, 2016; Vanyukov et al., 2012;
Wagner & Anthony, 2002) and self-medication hypothesis (Krause et al., 2002; Lawrence et al.,
2002; Levin et al., 2006). Specific attention was paid to the prevalence and epidemiology of ED
use behavior, associations between ED use behavior and substance use, associations between
ADHD and ED use behavior, and as concurrent influence of ED use and self-reported ADHD
symptoms on substance use behavior.
Summary of Findings
Study 1 found that current ED use was reported by roughly 20% of the participants in
both adolescent samples, while lifetime ED use prevalence was slightly higher among German
adolescents (55.9% vs. 35.1%). Those identified as Hispanic/Latino and multiethnic were more
likely to use EDs, while those identified as Asian were less likely to use them. ED use was cross-
sectionally associated with substance use variables (i.e., cigarettes, e-cigarettes, hookah, alcohol,
and marijuana) in both adolescent samples, even after controlling for covariates. Further, a
longitudinal analysis conducted among German adolescents demonstrated that ED use was
prospectively associated with increased risk of substance use initiation (cigarette, e-cigarette,
hookah, alcohol, marijuana). In fact, risk-taking did not moderate these associations except for
99
the model predicting marijuana use initiation: baseline ED use predicated marijuana use
initiation among the high risk-taking adolescent subgroup, but not in the low risk-taking
subgroup.
Study 2 found that ED use was cross-sectionally associated with self-reported ADHD
symptoms. Further, baseline self-reported ADHD symptoms predicted ED use initiation over a
subsequent 12-month period only among a subgroup of children aged 9-10 years. Additionally,
after controlling for baseline ADHD symptoms, baseline ED use predicted more frequent ADHD
symptoms at 12-months follow-up.
Study 3 found that ED consumption mediates the association between self-reported
ADHD symptoms and cigarette, e-cigarette, hookah, and alcohol use initiation. Specifically, self-
reported ADHD symptoms were positively associated with ED use behavior, which in turn was
associated with substance use initiation at 12-month follow-up. In sum, the research conducted in
this dissertation contributes to our understanding of ED use behavior and its associated substance
use consequences and have important implications for theory, future research methods, and ED
prevention and intervention programs, while also may be beneficial in improving the lives of
youth with ADHD symptoms.
Theoretical Implications
Theoretical background on ED use behavior is limited, nor is there a unique theory that
would explain ED use behavior among youth. This dissertation attempted to integrate multiple
theories and hypotheses in a single complex heuristic model (see Figure 1). Key findings from
dissertation studies support the use and validity of these theories and hypotheses while also
indicate that the proposed theoretical framework is a valuable tool for examining ED use
behavior and its associated consequences. Aligned with the problem behavior theory (Donovan
100
et al., 1991; Jessor, 1991; Jessor et al., 2003), studies 1 & 2 demonstrated that ED use is
associated with three interactive systems of psychosocial influence: personality (risk-taking,
depression symptoms), perceived environment (peer substance use, studying in schools other
than gymnasia), and behavior (substance use). Further, in line with the self-medication
hypothesis (Krause et al., 2002; Lawrence et al., 2002; Levin et al., 2006), study 2 suggested that
children (aged 9-10 years) with self-reported ADHD symptoms were more likely to initiate ED
use (possibly due to self-medication to cope with their symptoms) compared to children without
ADHD symptoms. Finally, aligned with the gateway hypothesis (Kandel, 1975; Kandel et al.,
1992; Nkansah-Amankra, 2020; Nkansah-Amankra & Minelli, 2016; Vanyukov et al., 2012;
Wagner & Anthony, 2002) and cross-sensitization hypothesis (Hellberg et al., 2019; Robinson et
al., 1985; Sussman, 2017; Temple, 2009), studies 1 & 3 support the assumption that repeated ED
consumption (among ADHD youth who cope with their symptoms by using EDs, as well as
among regular ED users) causes dysregulation in mesolimbic structures of the brain, which
progressively boosts motivation for EDs. As a result, ED sensitized individuals become more
prone to experience pathological motivation and have aggravated responses to other substances
that affect the same dopamine structures (i.e., cigarettes, e-cigarettes, hookah, alcohol); in other
words, repeated ED consumption acts as a “gateway” to use of other drugs. In sum, theories and
hypotheses integrated into the proposed framework provide a strong background and guidance
for these dissertation studies. Collectively results from these studies further support the use and
validity of these theories. At the same time, the proposed heuristic model provides a solid basis
for developing a new theory on multiple addictive behaviors.
101
Methodological Implications
A long-term, systematic assessment of ED use behavior is needed to better understand
how ED use changes over time. Specifically, it is vital to investigate ED use across the lifespan,
particularly at the age of 9-11 years, when most youths initiate using EDs (Costa et al., 2016;
Reid et al., 2015; Trapp et al., 2020) through emerging adulthood when ED use is the most
prevalent. Further, additional descriptive information of the most often consumed ED brand
should be examined. Given that EDs may not be the primary source of caffeine for children and
adolescents (Verster & Koenig, 2018), further research should account for other sources of
caffeine to rule out possible confounding.
These dissertation studies demonstrated longitudinal associations between ED use
behavior and substance; nonetheless, additional experimental animal studies utilizing a
sequential paradigm are warranted to test the causal validity of these associations. Further, some
evidence suggests that at least 30% of adolescents consume EDs at LAN parties (i.e., a gathering
of people to share a local area network [LAN] to play multiplayer video games), thus
investigating the associations between ED use, compulsive internet use and smartphone
addiction is important (Galimov et al., 2019). Further, studies suggest that some ED marketing
promotes other risky behaviors, including drug and alcohol use; hence, it is imperative to
examine whether exposure to these messages mediates the link between ED consumption and
substance use initiation (Harris & Munsell, 2015). Given that childhood and adolescence are
critical periods in the development of health-related behaviors thus, additional studies examining
the long-term health effects of EDs are warranted (Visram et al., 2016).
102
Programmatic Implications
Results from this dissertation may have important implications for prevention. Serval
intervention strategies could be used to reduce the risk of substance use initiation in adolescent
populations, particularly among youth with self-reported ADHD. These strategies could be
applied at the personal, interpersonal, community, and policy levels. For instance, on the
community level, implementing school-based prevention programs like Project Towards No
Drug Abuse (Sussman et al., 2002) may be beneficial. These or other similar programs could
implement interventions aimed at preventing the emergence of ED use behavior, which may help
reduce the risk for later substance use. Harm reduction strategies discouraging use in high-risk
contexts might emphasize education on EDs components and their adverse effects (Miller et al.,
2018).
Further, regular screening for ED use is warranted for youth with prevalent self-reported
ADHD symptoms. Promotion of healthy alternative outlets for sensation-seeking (i.e., going to
the gym) and teaching new “healthy” coping skills (i.e., mindfulness meditation) to manage
ADHD symptoms may lower the rates of substance use among children and adolescents with
ADHD. Given that some adolescent ED users endorse high-risk lifestyles (i.e., snowboarding),
targeting adolescents based on their sensation-seeking scores may be helpful. One study reported
that sensation seeking screening works moderately well at identifying adolescents at risk for
onset of binge drinking and established smoking (Sargent et al., 2010). Authors suggested that
these sensation-seeking scores may help identify high-risk adolescents that could be further
directed to the appropriate intervention program.
Adopting more strict age of sale policies may also be beneficial. In particular, U.S. and
Germany should consider adopting policies that would restrict the sale of EDs to minors. For
103
instance, EDs are not sold to adolescents under 15 years old in Sweden, while Denmark has
banned their sales completely (Oddy & O'sullivan, 2009). Moreover, a few studies suggested that
ED marketing, one of the leading factors associated with ED initiation (Galimov et al., 2019),
became more youth-oriented in recent years (Hammond & Reid, 2018; Harris, 2013; Harris &
Munsell, 2015; Stacey et al., 2017). Thus, new policies are warranted to ban the marketing of
EDs to youth under 21 years of age, including traditional and social media, sponsorships, and
other activities with youth audiences. ED manufacturers should be mandated to include the exact
caffeine and other ingredients content information on product packaging (including the amount
of caffeine in the entire container); additionally, all serious adverse events must be reported to
the FDA (or similar agency in Germany) (Harris & Munsell, 2015).
Interpersonal-level strategies may also be helpful. For instance, parents, school officials,
and healthcare providers need to be aware of signs of excessive EDs consumption and limit their
use by adolescents (Harris & Munsell, 2015). Other strategies aimed to minimize ED use among
youth could include targeted media campaigns encouraging parents to delay early initiation of
ED use habits. Promotion of healthy lifestyle alternatives, such as yoga or mindfulness
meditation, may be beneficial. Further, instruction in media literacy (i.e., making youth aware of
ED advertising influences) – which has shown great potential in tobacco-prevention research –
could reduce ED consumption among youth (Sussman et al., 1995).
OVERALL CONCLUSION
This dissertation sought to address gaps in ED use, substance use, and ADHD literature.
As was demonstrated, ED is much more prevalent than previously thought, particularly among
youth, and seems to be increasing exponentially. The observed link between ED use and
substance use initiation is of particular concern, especially among children and adolescents with
104
ADHD. It is crucial to continue research examining the factors that predict ED use behavior,
psychosocial correlates of ED use, and its associated consequences. The findings of these
dissertation studies highlight the urgent need for policy regulation and restriction of ED
consumption among children and adolescents. Healthcare providers, educators, and parents
should be familiar with EDs and the potential health consequences associated with their use.
These studies also yield important implications for youth with ADHD: reducing ED use among
ADHD children and adolescents may help curb the risk for later substance use initiation among
these individuals. Further steps aimed to increase public awareness and education about the
potential harms of ED are needed to form appropriate attitudes and beliefs among youth.
105
REFERENCES
Achenbach, T. M. (1991). Manual for the youth self-report and 1991 profile. Department of
Psychiatry, University of Vermont Burlington, VT.
AIM Market Insight. (2015). Global energy drinks market: Insights, market size, share, growth,
trends analysis and forecast to 2021. AIM Market Insight.
Al-Shaar, L., Vercammen, K., Lu, C., Richardson, S., Tamez, M., & Mattei, J. (2017). Health
effects and public health concerns of energy drink consumption in the United States: a
mini-review. Frontiers in public health, 5, 225.
Algorta, G. P., Dodd, A. L., Stringaris, A., & Youngstrom, E. A. (2016). Diagnostic efficiency of
the SDQ for parents to identify ADHD in the UK: a ROC analysis. European child &
adolescent psychiatry, 25(9), 949-957.
Ali, F., Rehman, H., Babayan, Z., Stapleton, D., & Joshi, D.-D. (2015). Energy drinks and their
adverse health effects: A systematic review of the current evidence. Postgraduate
Medicine, 127(3), 308-322. https://doi.org/10.1080/00325481.2015.1001712
American Academy of Pediatrics. (2011). Clinical Report—Sports drinks and energy drinks for
children and adolescents: Are they appropriate. Pediatrics, 127(6), 1182-1189.
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders
(DSM-5®). American Psychiatric Pub.
Azagba, S., Langille, D., & Asbridge, M. (2014). An emerging adolescent health risk:
Caffeinated energy drink consumption patterns among high school students. Preventive
Medicine, 62, 54-59. https://doi.org/10.1016/j.ypmed.2014.01.019
106
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 1173.
Barrington-Trimis, J. L., Bello, M. S., Liu, F., Leventhal, A. M., Kong, G., Mayer, M., Cruz, T.
B., Krishnan-Sarin, S., & McConnell, R. (2019). Ethnic differences in patterns of
cigarette and e-cigarette use over time among adolescents. Journal of Adolescent Health,
65(3), 359-365.
Barry, T. D., Lyman, R. D., & Klinger, L. G. (2002). Academic underachievement and attention-
deficit/hyperactivity disorder: The negative impact of symptom severity on school
performance. Journal of school psychology, 40(3), 259-283.
Bashir, D., Reed-Schrader, E., Olympia, R. P., Brady, J., Rivera, R., Serra, T., & Weber, C.
(2016). Clinical symptoms and adverse effects associated with energy drink consumption
in adolescents. Pediatric emergency care, 32(11), 751-755.
Bauman, A., & Phongsavan, P. (1999). Epidemiology of substance use in adolescence:
prevalence, trends and policy implications. Drug and Alcohol Dependence, 55(3), 187-
207.
Benner, A. D. (2011). The Transition to High School: Current Knowledge, Future Directions.
Educ Psychol Rev, 23(3), 299-328. https://doi.org/10.1007/s10648-011-9152-0
Biederman, J. (2005). Attention-deficit/hyperactivity disorder: a selective overview. Biological
psychiatry, 57(11), 1215-1220.
Biederman, J., Monuteaux, M. C., Mick, E., Wilens, T. E., Fontanella, J. A., Poetzl, K. M., Kirk,
T., Masse, J., & Faraone, S. V. (2006). Is cigarette smoking a gateway to alcohol and
107
illicit drug use disorders? A study of youths with and without attention deficit
hyperactivity disorder. Biological psychiatry, 59(3), 258-264.
Biederman, J., Wilens, T., Mick, E., Faraone, S. V., Weber, W., Curtis, S., Thornell, A., Pfister,
K., Jetton, J. G., & Soriano, J. (1997). Is ADHD a risk factor for psychoactive substance
use disorders? Findings from a four-year prospective follow-up study. Journal of the
American Academy of Child & Adolescent Psychiatry, 36(1), 21-29.
Bonnie, R. J., Stratton, K., & Kwan, L. Y. (2015). Public health implications of raising the
minimum age of legal access to tobacco products. National Academies Press
Washington, DC.
Brody, A. L., Mandelkern, M. A., Olmstead, R. E., Allen-Martinez, Z., Scheibal, D., Abrams, A.
L., Costello, M. R., Farahi, J., Saxena, S., & Monterosso, J. (2009). Ventral striatal
dopamine release in response to smoking a regular vs a denicotinized cigarette.
Neuropsychopharmacology, 34(2), 282-289.
Bunting, H., Baggett, A., & Grigor, J. (2013). Adolescent and young adult perceptions of
caffeinated energy drinks. A qualitative approach. Appetite, 65, 132-138.
Calvert, S. L. (2008). Children as consumers: Advertising and marketing. The future of children,
205-234.
Centers for Disease Control and Prevention. (2020). Age 21 Minimum Legal Drinking Age.
Retrieved November, 20 from https://www.cdc.gov/alcohol/fact-sheets/minimum-legal-
drinking-age.htm
Charach, A., Yeung, E., Climans, T., & Lillie, E. (2011). Childhood attention-
deficit/hyperactivity disorder and future substance use disorders: comparative meta-
108
analyses. Journal of the American Academy of Child & Adolescent Psychiatry, 50(1), 9-
21.
Chen, X., & Unger, J. B. (1999). Hazards of smoking initiation among Asian American and non-
Asian adolescents in California: a survival model analysis. Preventive Medicine, 28(6),
589-599.
Choi, H. J., Wolford-Clevenger, C., Brem, M. J., Elmquist, J., Stuart, G. L., Pasch, K. E., &
Temple, J. R. (2016). The temporal association between energy drink and alcohol use
among adolescents: A short communication. Drug and Alcohol Dependence, 158, 164-
166. https://doi.org/10.1016/j.drugalcdep.2015.11.009
Colder, C. R., Scalco, M., Trucco, E. M., Read, J. P., Lengua, L. J., Wieczorek, W. F., & Hawk,
L. W. (2013). Prospective associations of internalizing and externalizing problems and
their co-occurrence with early adolescent substance use. Journal of abnormal child
psychology, 41(4), 667-677.
Costa, B. M., Hayley, A., & Miller, P. (2014). Young adolescents’ perceptions, patterns, and
contexts of energy drink use. A focus group study. Appetite, 80, 183-189.
https://doi.org/10.1016/j.appet.2014.05.013
Costa, B. M., Hayley, A., & Miller, P. (2016). Adolescent energy drink consumption: An
Australian perspective. Appetite, 105, 638-642.
https://doi.org/10.1016/j.appet.2016.07.001
Covey, L. S., & Tam, D. (1990). Depressive mood, the single-parent home, and adolescent
cigarette smoking. Am J Public Health, 80(11), 1330-1333.
109
Cuffe, S. P., Moore, C. G., & McKeown, R. E. (2005). Prevalence and correlates of ADHD
symptoms in the national health interview survey. Journal of attention disorders, 9(2),
392-401.
De Alwis, D., Lynskey, M. T., Reiersen, A. M., & Agrawal, A. (2014). Attention-
deficit/hyperactivity disorder subtypes and substance use and use disorders in NESARC.
Addictive Behaviors, 39(8), 1278-1285.
de Bruijn, A., Tanghe, J., de Leeuw, R., Engels, R., Anderson, P., Beccaria, F., Bujalski, M.,
Celata, C., Gosselt, J., & Schreckenberg, D. (2016). European longitudinal study on the
relationship between adolescents’ alcohol marketing exposure and alcohol use. Addiction,
111(10), 1774-1783.
Di Rocco, J. R., During, A., Morelli, P. J., Heyden, M., & Biancaniello, T. A. (2011). Atrial
fibrillation in healthy adolescents after highly caffeinated beverage consumption: two
case reports. Journal of medical case reports, 5(1), 18.
Diamantopoulou, S., Henricsson, L., & Rydell, A.-M. (2005). ADHD symptoms and peer
relations of children in a community sample: examining associated problems, self-
perceptions, and gender differences. International Journal of Behavioral Development,
29(5), 388-398.
Dillon, P., Kelpin, S., Kendler, K., Thacker, L., Dick, D., & Svikis, D. (2019). Gender
Differences in Any-Source Caffeine and Energy Drink Use and Associated Adverse
Health Behaviors. Journal of caffeine and adenosine research, 9(1), 12-19.
Donovan, J. E., Jessor, R., & Costa, F. M. (1991). Adolescent health behavior and
conventionality-unconventionality: An extension of problem-behavior therapy. Health
Psychology, 10(1), 52.
110
East, K., Hitchman, S. C., Bakolis, I., Williams, S., Cheeseman, H., Arnott, D., & McNeill, A.
(2018). The association between smoking and electronic cigarette use in a cohort of
young people. Journal of Adolescent Health, 62(5), 539-547.
Egan, T. E., Dawson, A. E., & Wymbs, B. T. (2017). Substance use in undergraduate students
with histories of attention-deficit/hyperactivity disorder (ADHD): the role of impulsivity.
Substance use & misuse, 52(10), 1375-1386.
Ehlers, A., Marakis, G., Lampen, A., & Hirsch-Ernst, K. I. (2019). Risk assessment of energy
drinks with focus on cardiovascular parameters and energy drink consumption in Europe.
Food and Chemical Toxicology, 130, 109-121.
Emond, J. A., Gilbert-Diamond, D., Tanski, S. E., & Sargent, J. D. (2014). Energy Drink
Consumption and the Risk of Alcohol Use Disorder among a National Sample of
Adolescents and Young Adults. The Journal of Pediatrics, 165(6), 1194-1200.
https://doi.org/10.1016/j.jpeds.2014.08.050
Emond, J. A., Sargent, J. D., & Gilbert-Diamond, D. (2015). Patterns of Energy Drink
Advertising Over US Television Networks. Journal of Nutrition Education and Behavior,
47(2), 120-126.e121. https://doi.org/10.1016/j.jneb.2014.11.005
European Union Agency for Fundamental Rights. (2018). Purchasing and consuming alcohol.
Retrieved November, 20 from https://fra.europa.eu/en/publication/2017/mapping-
minimum-age-requirements/purchase-consumption-alcohol
Facts, P. (2013). Energy drinks and shots: US market trends. Rockville, MD: Packaged Facts.
Fairchild, A. J., & MacKinnon, D. P. (2009). A general model for testing mediation and
moderation effects. Prevention Science, 10(2), 87-99.
111
Faris, M. E. A. I. E., Jahrami, H., Al‐Hilali, M. M., Chehyber, N. J., Ali, S. O., Shahda, S. D., &
Obaid, R. S. (2017). Energy drink consumption is associated with reduced sleep quality
among college students: a cross‐sectional study. Nutrition & Dietetics, 74(3), 268-274.
Fisher, L. B., Miles, I., Austin, S., Camargo, C. A., Jr, & Colditz, G. A. (2007). Predictors of
initiation of alcohol use among us adolescents: Findings from a prospective cohort study.
Archives of Pediatrics & Adolescent Medicine, 161(10), 959-966.
https://doi.org/10.1001/archpedi.161.10.959
Fredholm, B. B., Bättig, K., Holmén, J., Nehlig, A., & Zvartau, E. E. (1999). Actions of caffeine
in the brain with special reference to factors that contribute to its widespread use.
Pharmacological reviews, 51(1), 83-133.
Froehlich, L., Martiny, S. E., & Deaux, K. (2019). A longitudinal investigation of the ethnic and
national identities of children with migration background in Germany. Social Psychology.
Galanti, M. R., Siliquini, R., Cuomo, L., Melero, J. C., Panella, M., & Faggiano, F. (2007).
Testing anonymous link procedures for follow-up of adolescents in a school-based trial:
The EU-DAP pilot study. Preventive Medicine, 44(2), 174-177.
https://doi.org/10.1016/j.ypmed.2006.07.019
Galimov, A., Hanewinkel, R., Hansen, J., Unger, J. B., Sussman, S., & Morgenstern, M. (2019).
Energy drink consumption among German adolescents: Prevalence, correlates, and
predictors of initiation. Appetite, 139, 172-179.
https://doi.org/https://doi.org/10.1016/j.appet.2019.04.016
Gallimberti, L., Buja, A., Chindamo, S., Vinelli, A., Lazzarin, G., Terraneo, A., Scafato, E., &
Baldo, V. (2013). Energy drink consumption in children and early adolescents [journal
112
article]. European Journal of Pediatrics, 172(10), 1335-1340.
https://doi.org/10.1007/s00431-013-2036-1
Galván, A. (2010). Adolescent development of the reward system [Review]. Frontiers in human
neuroscience, 4(6), 6. https://doi.org/10.3389/neuro.09.006.2010
Gardner, D. M., & Gerdes, A. C. (2015). A review of peer relationships and friendships in youth
with ADHD. Journal of attention disorders, 19(10), 844-855.
GmbH. (2016). Retail, Consumers, Advertising. Retrieved November 06, 2019, from
http://www.nielsen.com/content/dam/nielsenglobal/de/docs/Nielsen_Booklet_Retail-
Consumers-Advertising_2016.pdf
Goldman, L. S., Genel, M., Bezman, R. J., & Slanetz, P. J. (1998). Diagnosis and treatment of
attention-deficit/hyperactivity disorder in children and adolescents. JAMA, 279(14),
1100-1107.
Goodman, A., & Goodman, R. (2009). Strengths and difficulties questionnaire as a dimensional
measure of child mental health. Journal of the American Academy of Child & Adolescent
Psychiatry, 48(4), 400-403.
Goodman, E., Adler, N. E., Kawachi, I., Frazier, A. L., Huang, B., & Colditz, G. A. (2001).
Adolescents' Perceptions of Social Status: Development and Evaluation of a New
Indicator. J Pediatrics, 108(2), e31-e31. https://doi.org/10.1542/peds.108.2.e31
Goodman, R. (2001). Psychometric properties of the strengths and difficulties questionnaire.
Journal of the American Academy of Child & Adolescent Psychiatry, 40(11), 1337-1345.
Goodman, R., Ford, T., Simmons, H., Gatward, R., & Meltzer, H. (2003). Using the Strengths
and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a
community sample. International review of psychiatry, 15(1-2), 166-172.
113
Grasser, E. K., Yepuri, G., Dulloo, A. G., & Montani, J.-P. (2014). Cardio-and cerebrovascular
responses to the energy drink Red Bull in young adults: a randomized cross-over study.
European journal of nutrition, 53(7), 1561-1571.
Grichnik, D. (2008). Risky choices in new venture decisions—experimental evidence from
Germany and the United States. Journal of International Entrepreneurship, 6(1), 22-47.
Gunja, N., & Brown, J. A. (2012). Energy drinks: health risks and toxicity. Med J Aust, 196(1),
46-49.
Hall, C. L., Guo, B., Valentine, A. Z., Groom, M. J., Daley, D., Sayal, K., & Hollis, C. (2019).
The validity of the Strengths and Difficulties Questionnaire (SDQ) for children with
ADHD symptoms. PloS one, 14(6), e0218518.
Hamilton, H. A., Boak, A., Ilie, G., & Mann, R. E. (2013). Energy drink consumption and
associations with demographic characteristics, drug use and injury among adolescents.
Can J Public Health, 104(7), e496-501.
Hammond, D., & Reid, J. L. (2018). Exposure and perceptions of marketing for caffeinated
energy drinks among young Canadians. Public Health Nutrition, 21(3), 535-542.
Hammond, D., Reid, J. L., & Zukowski, S. (2018). Adverse effects of caffeinated energy drinks
among youth and young adults in Canada: a web-based survey. CMAJ open, 6(1), E19.
Hampson, S. E., Andrews, J. A., & Barckley, M. (2008). Childhood predictors of adolescent
marijuana use: Early sensation-seeking, deviant peer affiliation, and social images.
Addictive Behaviors, 33(9), 1140-1147.
Hanewinkel, R., Isensee, B., Sargent, J. D., & Morgenstern, M. (2011). Cigarette advertising and
teen smoking initiation. Pediatrics, peds. 2010-2934.
114
Hansen, J., Hanewinkel, R., & Morgenstern, M. (2018). Electronic cigarette marketing and
smoking behaviour in adolescence: a cross-sectional study. ERJ open research, 4(4),
00155-02018.
Harris, J. L. (2013). Energy drink marketing to teens: 2010 to 2013. Retrieved November 06,
2019, from
http://www.uconnruddcenter.org/files/Pdfs/Energy_Drink_Marketing_Teens.pdf
Harris, J. L., & Munsell, C. R. (2015). Energy drinks and adolescents: what’s the harm?
Nutrition Reviews, 73(4), 247-257. https://doi.org/10.1093/nutrit/nuu061
Harris, J. L., Schwartz, M., Brownell, K. D., Javadizadeh, J., & Weinberg, M. (2011). Evaluating
sugary drink nutrition and marketing to youth. New Haven, CT: Yale Rudd Center For
Food Policy and Obesity.
Hellberg, S. N., Russell, T. I., & Robinson, M. J. (2019). Cued for risk: Evidence for an
incentive sensitization framework to explain the interplay between stress and anxiety,
substance abuse, and reward uncertainty in disordered gambling behavior. Cognitive,
Affective, & Behavioral Neuroscience, 19(3), 737-758.
Hofstede, G. (1994). Management scientists are human. Management science, 40(1), 4-13.
Hughes, M., Harrison, E., Peel, N., Stutchfield, B., McNally, S., Beattie, C., & Wigmore, S.
(2015). Randomized clinical trial of perioperative nerve block and continuous local
anaesthetic infiltration via wound catheter versus epidural analgesia in open liver
resection (LIVER 2 trial). British Journal of Surgery, 102(13), 1619-1628.
Jensen, P. S., Hinshaw, S. P., Kraemer, H. C., Lenora, N., Newcorn, J. H., Abikoff, H. B.,
March, J. S., Arnold, L. E., Cantwell, D. P., & Conners, C. K. (2001). ADHD
115
comorbidity findings from the MTA study: comparing comorbid subgroups. Journal of
the American Academy of Child & Adolescent Psychiatry, 40(2), 147-158.
Jessor, R. (1991). Risk behavior in adolescence: A psychosocial framework for understanding
and action. Journal of Adolescent Health, 12(8), 597-605. https://doi.org/10.1016/1054-
139X(91)90007-K
Jessor, R., Turbin, M. S., Costa, F. M., Dong, Q., Zhang, H., & Wang, C. (2003). Adolescent
problem behavior in China and the United States: A cross‐national study of psychosocial
protective factors. Journal of Research on adolescence, 13(3), 329-360.
Johnston, L., Miech, R., O'Malley, P., Bachman, J., Schulenberg, J., & Patrick, M. (2020).
Monitoring the Future national survey results on drug use, 1975-2019: Overview, key
findings on adolescent drug use.
Kandel, D. (1975). Stages in adolescent involvement in drug use. Science, 190(4217), 912-914.
Kandel, D. B., Yamaguchi, K., & Chen, K. (1992). Stages of progression in drug involvement
from adolescence to adulthood: further evidence for the gateway theory. Journal of
studies on alcohol, 53(5), 447-457.
Kann, L., McManus, T., Harris, W. A., Shanklin, S. L., Flint, K. H., Queen, B., Lowry, R.,
Chyen, D., Whittle, L., & Thornton, J. (2018). Youth risk behavior surveillance—United
States, 2017. MMWR Surveillance Summaries, 67(8), 1.
Katusic, S. K., Barbaresi, W. J., Colligan, R. C., Weaver, A. L., Leibson, C. L., & Jacobsen, S. J.
(2005). Psychostimulant Treatment and Risk for Substance Abuse Among Young Adults
with a History of Attention-Deficit/Hyperactivity Disorder: APopulation-Based, Birth
Cohort Study. Journal of Child & Adolescent Psychopharmacology, 15(5), 764-776.
116
Kaur, S., Christian, H., Cooper, M. N., Francis, J., Allen, K., & Trapp, G. Consumption of
energy drinks is associated with depression, anxiety, and stress in young adult males:
Evidence from a longitudinal cohort study. Depression and Anxiety.
Kessler, R. C., Avenevoli, S., McLaughlin, K., Green, J. G., Lakoma, M., Petukhova, M., Pine,
D., Sampson, N., Zaslavsky, A., & Merikangas, K. R. (2012). Lifetime comorbidity of
DSM-IV disorders in the NCS-R adolescent supplement (NCS-A). Psychological
medicine, 42(9), 1997.
Koivusilta, L., Kuoppamäki, H., & Rimpelä, A. (2016). Energy drink consumption, health
complaints and late bedtime among young adolescents [journal article]. International
Journal of Public Health, 61(3), 299-306. https://doi.org/10.1007/s00038-016-0797-9
Konofal, E., Lecendreux, M., & Cortese, S. (2010). Sleep and ADHD. Sleep medicine, 11(7),
652-658.
Kozlowski, L. T., & Warner, K. E. (2017). Adolescents and e-cigarettes: Objects of concern may
appear larger than they are. Drug and Alcohol Dependence.
Krause, K.-H., Dresel, S. H., Krause, J., Kung, H. F., Tatsch, K., & Ackenheil, M. (2002).
Stimulant-like action of nicotine on striatal dopamine transporter in the brain of adults
with attention deficit hyperactivity disorder. International Journal of
Neuropsychopharmacology, 5(2), 111-113.
Kumar, G., Park, S., & Onufrak, S. (2015). Perceptions about energy drinks are associated with
energy drink intake among US youth. American Journal of Health Promotion, 29(4),
238-244.
Kunkel, D., & Castonguay, J. (2012). Children and advertising: Content, comprehension, and
consequences.
117
Lafreniere, K., Menna, R., & Cramer, K. M. (2013). Rebelliousness, effortful control, and risky
behavior: Metamotivational and temperamental predictors of risk-taking in older
adolescents. Journal of Motivation, Emotion, and Personality, 1, 17.
Larson, N., DeWolfe, J., Story, M., & Neumark-Sztainer, D. (2014). Adolescent Consumption of
Sports and Energy Drinks: Linkages to Higher Physical Activity, Unhealthy Beverage
Patterns, Cigarette Smoking, and Screen Media Use. Journal of Nutrition Education and
Behavior, 46(3), 181-187. https://doi.org/10.1016/j.jneb.2014.02.008
Larson, N., Laska, M. N., Story, M., & Neumark-Sztainer, D. (2015). Sports and energy drink
consumption are linked to health-risk behaviours among young adults. Public Health
Nutrition, 18(15), 2794-2803.
Lawrence, N. S., Ross, T. J., & Stein, E. A. (2002). Cognitive mechanisms of nicotine on visual
attention. Neuron, 36(3), 539-548.
Leal, W. E., & Jackson, D. B. (2018). Energy drinks and escalation in drug use severity: An
emergent hazard to adolescent health. Preventive Medicine, 111, 391-396.
https://doi.org/https://doi.org/10.1016/j.ypmed.2017.11.033
Leventhal, A. M., Strong, D. R., Kirkpatrick, M. G., Unger, J. B., Sussman, S., Riggs, N. R.,
Stone, M. D., Khoddam, R., Samet, J. M., & Audrain-McGovern, J. (2015). Association
of electronic cigarette use with initiation of combustible tobacco product smoking in
early adolescence. JAMA, 314(7), 700-707.
Levin, E. D., McClernon, F. J., & Rezvani, A. H. (2006). Nicotinic effects on cognitive function:
behavioral characterization, pharmacological specification, and anatomic localization.
Psychopharmacology, 184(3-4), 523-539.
118
Li, Y., & Lerner, R. M. (2011). Trajectories of school engagement during adolescence:
Implications for grades, depression, delinquency, and substance use. Developmental
Psychology, 47(1), 233-247. https://doi.org/10.1037/a0021307
Link, T. C. (2008). Adolescent substance use in Germany and the United States: A cross-cultural
test of the applicability and generalizability of theoretical indicators. European Journal of
Criminology, 5(4), 453-480.
Malinauskas, B. M., Aeby, V. G., Overton, R. F., Carpenter-Aeby, T., & Barber-Heidal, K.
(2007). A survey of energy drink consumption patterns among college students. Nutrition
Journal, 6(1), 35. https://doi.org/10.1186/1475-2891-6-35
Markey, E. J., J. Durbin R.J, R., & Blumenthal, R. (2014). What's all the buzz about? A survey of
popular energy drinks finds inconsistent labeling, questionable ingredients and targeted
marketing to adolescents.
Marmorstein, N. R. (2016). Energy drink and coffee consumption and psychopathology
symptoms among early adolescents: cross-sectional and longitudinal associations.
Journal of caffeine research, 6(2), 64-72.
Martinelli, T., Candel, M. J., de Vries, H., Talhout, R., Knapen, V., van Schayck, C. P., &
Nagelhout, G. E. (2021). Exploring the gateway hypothesis of e-cigarettes and tobacco: a
prospective replication study among adolescents in the Netherlands and Flanders. Tob
Control.
Masengo, L., Sampasa-Kanyinga, H., Chaput, J.-P., Hamilton, H. A., & Colman, I. (2020).
Energy drink consumption, psychological distress, and suicidality among middle and
high school students. Journal of affective disorders.
119
McCusker, R. R., Goldberger, B. A., & Cone, E. J. (2006). Caffeine Content of Energy Drinks,
Carbonated Sodas, and Other Beverages. Journal of Analytical Toxicology, 30(2), 112-
114. https://doi.org/10.1093/jat/30.2.112
McKetin, R., Coen, A., & Kaye, S. (2015). A comprehensive review of the effects of mixing
caffeinated energy drinks with alcohol. Drug and Alcohol Dependence, 151, 15-30.
Meredith, S. E., Juliano, L. M., Hughes, J. R., & Griffiths, R. R. (2013). Caffeine use disorder: a
comprehensive review and research agenda. Journal of caffeine research, 3(3), 114-130.
Miller, K. E. (2008a). Energy Drinks, Race, and Problem Behaviors Among College Students.
Journal of Adolescent Health, 43(5), 490-497.
https://doi.org/10.1016/j.jadohealth.2008.03.003
Miller, K. E. (2008b). Wired: Energy Drinks, Jock Identity, Masculine Norms, and Risk Taking.
Journal of American College Health, 56(5), 481-490.
https://doi.org/10.3200/JACH.56.5.481-490
Miller, K. E., Dermen, K. H., & Lucke, J. F. (2018). Caffeinated energy drink use by US
adolescents aged 13–17: A national profile. Psychology of Addictive Behaviors, 32(6),
647.
Miyake, E. R., & Marmorstein, N. R. (2015). Energy drink consumption and later alcohol use
among early adolescents. Addictive Behaviors, 43, 60-65.
https://doi.org/10.1016/j.addbeh.2014.12.009
Morgenstern, M., Sargent, J. D., Isensee, B., & Hanewinkel, R. (2013). From never to daily
smoking in 30 months: the predictive value of tobacco and non-tobacco advertising
exposure. BMJ Open, 3(6). https://doi.org/10.1136/bmjopen-2013-002907
120
Moss, S. L., Santaella‐Tenorio, J., Mauro, P. M., Keyes, K. M., & Martins, S. S. (2019). Changes
over time in marijuana use, deviant behavior and preference for risky behavior among US
adolescents from 2002 to 2014: testing the moderating effect of gender and age.
Addiction, 114(4), 674-686.
Muris, P., Meesters, C., & van den Berg, F. (2003). The strengths and difficulties questionnaire
(SDQ). European child & adolescent psychiatry, 12(1), 1-8.
Nall, A. H., Shakhmantsir, I., Cichewicz, K., Birman, S., Hirsh, J., & Sehgal, A. (2016). Caffeine
promotes wakefulness via dopamine signaling in Drosophila. Scientific reports, 6, 20938.
Nehlig, A., Daval, J.-L., & Debry, G. (1992). Caffeine and the central nervous system:
mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain
Research Reviews, 17(2), 139-170.
Nkansah-Amankra, S. (2020). Revisiting the association between “gateway hypothesis” of early
drug use and drug use progression: A cohort analysis of peer influences on drug use
progression among a population cohort. Substance use & misuse, 55(6), 998-1007.
Nkansah-Amankra, S., & Minelli, M. (2016). “Gateway hypothesis” and early drug use:
Additional findings from tracking a population-based sample of adolescents to adulthood.
Preventive medicine reports, 4, 134-141.
Nowak, D., & Jasionowski, A. (2015). Analysis of the consumption of caffeinated energy drinks
among Polish adolescents. International journal of environmental research and public
health, 12(7), 7910-7921.
Nuyts, P. A., Kuipers, M. A., Willemsen, M. C., & Kunst, A. E. (2020). An increase in the
tobacco age-of-sale to 21: for debate in Europe. Nicotine and Tobacco Research, 22(7),
1247-1249.
121
Nyarko, K. A., Grosse, S. D., Danielson, M. L., Holbrook, J. R., Visser, S. N., & Shapira, S. K.
(2017). Treated prevalence of attention-deficit/hyperactivity disorder increased from
2009 to 2015 among school-aged children and adolescents in the United States. Journal
of Child and Adolescent Psychopharmacology, 27(8), 731-734.
Oddy, W., & O'sullivan, T. (2009). Energy drinks for children and adolescents. BMJ: British
Medical Journal (Online), 339.
Ostroumov, A., Thomas, A. M., Dani, J. A., & Doyon, W. M. (2015). Cigarettes and alcohol:
The influence of nicotine on operant alcohol self-administration and the mesolimbic
dopamine system. Biochemical pharmacology, 97(4), 550-557.
Pan, P. Y., & Yeh, C. B. (2017). Impact of depressive/anxiety symptoms on the quality of life of
adolescents with ADHD: a community-based 1-year prospective follow-up study.
European child & adolescent psychiatry, 26(6), 659-667.
Park, S., Blanck, H. M., Sherry, B., Brener, N., & O'Toole, T. (2012). Factors associated with
sugar-sweetened beverage intake among United States high school students. The Journal
of nutrition, 142(2), 306-312.
Park, S., Lee, Y., & Lee, J. H. (2016). Association between energy drink intake, sleep, stress, and
suicidality in Korean adolescents: energy drink use in isolation or in combination with
junk food consumption. Nutrition Journal, 15(1), 87.
Pennington, N., Johnson, M., Delaney, E., & Blankenship, M. B. (2010). Energy drinks: a new
health hazard for adolescents. The Journal of School Nursing, 26(5), 352-359.
Pollak, Y., Dekkers, T. J., Shoham, R., & Huizenga, H. M. (2019). Risk-taking behavior in
attention deficit/hyperactivity disorder (ADHD): A review of potential underlying
mechanisms and of interventions. Current psychiatry reports, 21(5), 1-11.
122
Pomeranz, J. L., Munsell, C. R., & Harris, J. L. (2013). Energy drinks: an emerging public health
hazard for youth. Journal of public health policy, 34(2), 254-271.
Potter, A. S., & Newhouse, P. A. (2008). Acute nicotine improves cognitive deficits in young
adults with attention-deficit/hyperactivity disorder. Pharmacology Biochemistry and
Behavior, 88(4), 407-417.
Poulos, N. S., & Pasch, K. E. (2015). Energy drink consumption is associated with unhealthy
dietary behaviours among college youth. Perspectives in Public Health, 135(6), 316-321.
https://doi.org/10.1177/1757913914565388
Rapport, M. D., Scanlan, S. W., & Denney, C. B. (1999). Attention-deficit/hyperactivity disorder
and scholastic achievement: A model of dual developmental pathways. The Journal of
Child Psychology and Psychiatry and Allied Disciplines, 40(8), 1169-1183.
Reid, J. L., Hammond, D., McCrory, C., Dubin, J. A., & Leatherdale, S. T. (2015). Use of
caffeinated energy drinks among secondary school students in Ontario: Prevalence and
correlates of using energy drinks and mixing with alcohol. Can J Public Health, 106(3),
101-108.
Reid, J. L., McCrory, C., White, C. M., Martineau, C., Vanderkooy, P., Fenton, N., & Hammond,
D. (2017). Consumption of Caffeinated Energy Drinks Among Youth and Young Adults
in Canada. Preventive Medicine Reports, 5, 65-70.
https://doi.org/10.1016/j.pmedr.2016.11.012
Reissig, C. J., Strain, E. C., & Griffiths, R. R. (2009). Caffeinated energy drinks—A growing
problem. Drug and Alcohol Dependence, 99(1), 1-10.
https://doi.org/10.1016/j.drugalcdep.2008.08.001
123
Retzbach, E. P., Dholakia, P. H., & Duncan-Vaidya, E. A. (2014). The effect of daily caffeine
exposure on lever-pressing for sucrose and c-Fos expression in the nucleus accumbens in
the rat. J Physiology behavior, 135, 1-6.
Richards, G., & Smith, A. P. (2016a). Breakfast and Energy Drink Consumption in Secondary
School Children: Breakfast Omission, in Isolation or in Combination with Frequent
Energy Drink Use, is Associated with Stress, Anxiety, and Depression Cross-Sectionally,
but not at 6-Month Follow-Up. Frontiers in Psychology, 7(106).
https://doi.org/10.3389/fpsyg.2016.00106
Richards, G., & Smith, A. P. (2016b). A review of energy drinks and mental health, with a focus
on stress, anxiety, and depression. Journal of Caffeine Research, 6(2), 49-63.
Robinson, T. E., Angus, A. L., & Becker, J. B. (1985). Sensitization to stress: the enduring
effects of prior stress on amphetamine-induced rotational behavior. Life sciences, 37(11),
1039-1042.
Robinson, T. E., & Berridge, K. C. (1993). The neural basis of drug craving: an incentive-
sensitization theory of addiction. Brain Research Reviews, 18(3), 247-291.
Robinson, T. E., & Berridge, K. C. (2008). The incentive sensitization theory of addiction: some
current issues. Philosophical Transactions of the Royal Society B: Biological Sciences,
363(1507), 3137-3146.
Rogers, M. A., Wiener, J., Marton, I., & Tannock, R. (2009). Parental involvement in children's
learning: Comparing parents of children with and without Attention-Deficit/Hyperactivity
Disorder (ADHD). Journal of school psychology, 47(3), 167-185.
124
Roy, A., & Deshmukh, R. (2019). Energy Drinks Market by Type (Alcoholic and Nonalcoholic)
and End User (kids, Adults, and Teenagers): Global Opportunity Analysis and Industry
Forecast, 2019–2026. A. M. Research.
Sampasa-Kanyinga, H., Masengo, L., Hamilton, H. A., & Chaput, J.-P. (2020). Energy drink
consumption and substance use among middle and high school students. International
journal of environmental research and public health, 17(9), 3110.
Sanchis-Gomar, F., Pareja-Galeano, H., Cervellin, G., Lippi, G., & Earnest, C. P. (2015). Energy
drink overconsumption in adolescents: implications for arrhythmias and other
cardiovascular events. Canadian Journal of Cardiology, 31(5), 572-575.
Sargent, J. D., Tanski, S., Stoolmiller, M., & Hanewinkel, R. (2010). Using sensation seeking to
target adolescents for substance use interventions. Addiction, 105(3), 506-514.
Schenk, L., Ellert, U., & Neuhauser, H. (2007). Children and adolescents in Germany with a
migration background. Methodical aspects in the German Health Interview and
Examination Survey for Children and Adolescents (KiGGS). Bundesgesundheitsblatt,
Gesundheitsforschung, Gesundheitsschutz, 50(5-6), 590-599.
Schneider, M. B., & Benjamin, H. J. (2011). Sports drinks and energy drinks for children and
adolescents: are they appropriate? Pediatrics, 127(6), 1182-1189.
Schneider, S. K., Buka, S. L., Dash, K., Winickoff, J. P., & O'Donnell, L. (2016). Community
reductions in youth smoking after raising the minimum tobacco sales age to 21. Tob
Control, 25(3), 355-359.
Schwartz, D. L., Gilstad-Hayden, K., Carroll-Scott, A., Grilo, S. A., McCaslin, C., Schwartz, M.,
& Ickovics, J. R. (2015). Energy drinks and youth self-reported hyperactivity/inattention
symptoms. Academic pediatrics, 15(3), 297-304.
125
Scott, M. J., El-Hassan, M., & Khan, A. A. (2011). Myocardial infarction in a young adult
following the consumption of a caffeinated energy drink. Case Reports, 2011,
bcr0220113854.
Seidel, A.-K., Morgenstern, M., Galimov, A., Pedersen, A., Isensee, B., Goecke, M., &
Hanewinkel, R. (2021). Use of electronic cigarettes as a predictor of cannabis
experimentation: A longitudinal study among German youth. Nicotine & Tobacco
Research.
Seifert, S. M., Schaechter, J. L., Hershorin, E. R., & Lipshultz, S. E. (2011). Health effects of
energy drinks on children, adolescents, and young adults. Pediatrics, peds. 2009-3592.
Sharif, I., & Sargent, J. D. (2006). Association Between Television, Movie, and Video Game
Exposure and School Performance. Pediatrics, 118(4), e1061-e1070.
https://doi.org/10.1542/peds.2005-2854
Silverman, K., Evans, S. M., Strain, E. C., & Griffiths, R. R. (1992). Withdrawal syndrome after
the double-blind cessation of caffeine consumption. New England Journal of Medicine,
327(16), 1109-1114.
Solinas, M., Ferre, S., You, Z.-B., Karcz-Kubicha, M., Popoli, P., & Goldberg, S. R. (2002).
Caffeine induces dopamine and glutamate release in the shell of the nucleus accumbens.
Journal of Neuroscience, 22(15), 6321-6324.
Somers, K. R., & Svatikova, A. (2020). Cardiovascular and Autonomic Responses to Energy
Drinks—Clinical Implications. Journal of Clinical Medicine, 9(2), 431.
Stacey, N., van Walbeek, C., Maboshe, M., Tugendhaft, A., & Hofman, K. (2017). Energy drink
consumption and marketing in South Africa. Preventive Medicine, 105, S32-S36.
126
Steinberg, L. (2008). A Social Neuroscience Perspective on Adolescent Risk-Taking. Dev Rev,
28(1), 78-106. https://doi.org/10.1016/j.dr.2007.08.002
Stephenson, M. T., Hoyle, R. H., Palmgreen, P., & Slater, M. D. (2003). Brief measures of
sensation seeking for screening and large-scale surveys. Drug and Alcohol Dependence,
72(3), 279-286. https://doi.org/10.1016/j.drugalcdep.2003.08.003
Stokes, A., Wilson, A. E., Lundberg, D. J., Xie, W., Berry, K. M., Fetterman, J. L., Harlow, A.
F., Cozier, Y. C., Barrington-Trimis, J. L., & Sterling, K. L. (2020). Racial/Ethnic
Differences in Associations of Noncigarette Tobacco Product Use with Subsequent
Initiation of Cigarettes in US Youths. Nicotine & Tobacco Research.
Strine, T. W., Lesesne, C. A., Okoro, C. A., McGuire, L. C., Chapman, D. P., Balluz, L. S., &
Mokdad, A. H. (2006). Emotional and behavioral difficulties and impairments in
everyday functioning among children with a history of attention-deficit/hyperactivity
disorder. Preventing chronic disease, 3(2).
Sussman, S. (2017). Substance and behavioral addictions: Concepts, causes, and cures.
Cambridge University Press.
Sussman, S., & Ames, S. L. (2008). Drug abuse: Concepts, prevention, and cessation.
Cambridge University Press.
Sussman, S., Dent, C. W., & Stacy, A. W. (2002). Project Towards No Drug Abuse: A review of
the findings and future directions. American Journal of Health Behavior, 26(5), 354-365.
Swanson, J. M., Kinsbourne, M., Nigg, J., Lanphear, B., Stefanatos, G. A., Volkow, N., Taylor,
E., Casey, B., Castellanos, F. X., & Wadhwa, P. D. (2007). Etiologic subtypes of
attention-deficit/hyperactivity disorder: brain imaging, molecular genetic and
127
environmental factors and the dopamine hypothesis. Neuropsychology review, 17(1), 39-
59.
Temple, J. L. (2009). Caffeine use in children: What we know, what we have left to learn, and
why we should worry. Neuroscience & Biobehavioral Reviews, 33(6), 793-806.
https://doi.org/https://doi.org/10.1016/j.neubiorev.2009.01.001
Terry-McElrath, Y. M., O’Malley, P. M., & Johnston, L. D. (2014). Energy drinks, soft drinks,
and substance use among US secondary school students. Journal of addiction medicine,
8(1), 6.
Thapar, A., Cooper, M., Eyre, O., & Langley, K. (2013). Practitioner review: what have we
learnt about the causes of ADHD? Journal of Child Psychology and Psychiatry, 54(1), 3-
16.
Thomas, R., Sanders, S., Doust, J., Beller, E., & Glasziou, P. (2015). Prevalence of attention-
deficit/hyperactivity disorder: a systematic review and meta-analysis. Pediatrics, 135(4),
e994-e1001.
Trapp, Hurworth, M., Christian, H., Bromberg, M., Howard, J., McStay, C., Ambrosini, G.,
Martin, K., Harray, A., & Cross, D. (2020). Prevalence and pattern of energy drink intake
among Australian adolescents. Journal of Human Nutrition and Dietetics.
Trapp, G., Hurworth, M., Jacoby, P., Maddison, K., Allen, K., Martin, K., Christian, H.,
Ambrosini, G. L., Oddy, W., & Eastwood, P. R. (2020). Energy drink intake is associated
with insomnia and decreased daytime functioning in young adult females. Public Health
Nutrition, 1-10.
Triandis, H. C., McCusker, C., & Hui, C. H. (1990). Multimethod probes of individualism and
collectivism. Journal of Personality and Social Psychology, 59(5), 1006.
128
Tripp, G., & Wickens, J. R. (2009). Neurobiology of ADHD. Neuropharmacology, 57(7-8), 579-
589.
Tyas, S. L., & Pederson, L. L. (1998). Psychosocial factors related to adolescent smoking: a
critical review of the literature. Tob Control, 7(4), 409-420.
Unger, J. B., Ritt-Olson, A., Teran, L., Huang, T., Hoffman, B. R., & Palmer, P. (2002). Cultural
values and substance use in a multiethnic sample of California adolescents. Addiction
Research & Theory, 10(3), 257-279.
Unger, J. B., Rohrbach, L. A., Cruz, T. B., Baezconde-Garbanati, L., Palmer, P. H., Johnson, C.
A., & Howard, K. A. (2001). Ethnic variation in peer influences on adolescent smoking.
Nicotine & Tobacco Research, 3(2), 167-176.
US Food and Drug Administration. (2020). Tobacco 21. Retrieved November, 02 from
https://www.fda.gov/tobacco-products/retail-sales-tobacco-products/tobacco-21
Usman, A., & Jawaid, A. (2012). Hypertension in a young boy: an energy drink effect. BMC
research notes, 5(1), 1-3.
Vanyukov, M. M., Tarter, R. E., Kirillova, G. P., Kirisci, L., Reynolds, M. D., Kreek, M. J.,
Conway, K. P., Maher, B. S., Iacono, W. G., & Bierut, L. (2012). Common liability to
addiction and “gateway hypothesis”: theoretical, empirical and evolutionary perspective.
Drug and alcohol dependence, 123, S3-S17.
Vercammen, K. A., Koma, J. W., & Bleich, S. N. (2019). Trends in energy drink consumption
among US adolescents and adults, 2003–2016. American journal of preventive medicine,
56(6), 827-833.
129
Verster, J. C., & Koenig, J. (2018). Caffeine intake and its sources: A review of national
representative studies. Critical Reviews in Food Science and Nutrition, 58(8), 1250-1259.
https://doi.org/10.1080/10408398.2016.1247252
Visram, S., Cheetham, M., Riby, D. M., Crossley, S. J., & Lake, A. A. (2016). Consumption of
energy drinks by children and young people: a rapid review examining evidence of
physical effects and consumer attitudes. BMJ Open, 6(10).
Visram, S., Crossley, S. J., Cheetham, M., & Lake, A. (2017). Children and young people’s
perceptions of energy drinks: A qualitative study. PloS one, 12(11), e0188668.
Visser, S. N., Bitsko, R. H., Danielson, M. L., Perou, R., & Blumberg, S. J. (2010). Increasing
prevalence of parent-reported attention-deficit/hyperactivity disorder among children—
United States, 2003 and 2007. Morbidity and mortality weekly report, 59(44), 1439-1443.
Volkow, N. D., Wang, G.-J., Kollins, S. H., Wigal, T. L., Newcorn, J. H., Telang, F., Fowler, J.
S., Zhu, W., Logan, J., & Ma, Y. (2009). Evaluating dopamine reward pathway in
ADHD: clinical implications. JAMA, 302(10), 1084-1091.
Wagenaar, A. C., & Toomey, T. L. (2002). Effects of minimum drinking age laws: review and
analyses of the literature from 1960 to 2000. Journal of Studies on Alcohol,
supplement(14), 206-225.
Wagner, F. A., & Anthony, J. C. (2002). Into the world of illegal drug use: exposure opportunity
and other mechanisms linking the use of alcohol, tobacco, marijuana, and cocaine.
American journal of epidemiology, 155(10), 918-925.
Walker, A. L., Peters, P. H., de Rooij, S. R., Henrichs, J., Witteveen, A. B., Verhoeven, C. J.,
Vrijkotte, T. G., & de Jonge, A. (2020). The Long-Term Impact of Maternal Anxiety and
130
Depression Postpartum and in Early Childhood on Child and Paternal Mental Health at
11–12 Years Follow-Up. Frontiers in Psychiatry, 11, 918.
Weinstein, N. D. (1982). Unrealistic optimism about susceptibility to health problems. Journal of
behavioral medicine, 5(4), 441-460.
White, I. R., Royston, P., & Wood, A. M. (2011). Multiple imputation using chained equations:
Issues and guidance for practice. Statistics in Medicine, 30(4), 377-399.
https://doi.org/10.1002/sim.4067
Wilens, T. E., Martelon, M., Joshi, G., Bateman, C., Fried, R., Petty, C., & Biederman, J. (2011).
Does ADHD predict substance-use disorders? A 10-year follow-up study of young adults
with ADHD. Journal of the American Academy of Child & Adolescent Psychiatry, 50(6),
543-553.
Wills, T. A., Vaccaro, D., & McNamara, G. (1994). Novelty seeking, risk taking, and related
constructs as predictors of adolescent substance use: An application of Cloninger's
theory. Journal of Substance Abuse, 6(1), 1-20.
https://doi.org/https://doi.org/10.1016/S0899-3289(94)90039-6
Wilson, R. E., Kado, H. S., Samson, R., & Miller, A. B. (2012). A case of caffeine-induced
coronary artery vasospasm of a 17-year-old male. Cardiovascular toxicology, 12(2), 175-
179.
Woolsey, C. L., Barnes, L. B., Jacobson, B. H., Kensinger, W. S., Barry, A. E., Beck, N. C.,
Resnik, A. G., & Evans, M. W. (2014). Frequency of Energy Drink Use Predicts Illicit
Prescription Stimulant Use. Substance Abuse, 35(1), 96-103.
https://doi.org/10.1080/08897077.2013.810561
131
Yen, J.-Y., Ko, C.-H., Yen, C.-F., Wu, H.-Y., & Yang, M.-J. (2007). The comorbid psychiatric
symptoms of Internet addiction: attention deficit and hyperactivity disorder (ADHD),
depression, social phobia, and hostility. Journal of Adolescent Health, 41(1), 93-98.
Zhang, W., O’Brien, N., Forrest, J. I., Salters, K. A., Patterson, T. L., Montaner, J. S., Hogg, R.
S., & Lima, V. D. (2012). Validating a shortened depression scale (10 item CES-D)
among HIV-positive people in British Columbia, Canada. PloS one, 7(7).
Zucconi, S., Volpato, C., Adinolfi, F., Gandini, E., Gentile, E., Loi, A., & Fioriti, L. (2013).
Gathering consumption data on specific consumer groups of energy drinks. EFSA
Supporting Publications, 10(3), 394E. https://doi.org/10.2903/sp.efsa.2013.EN-394
Abstract (if available)
Abstract
Energy drinks (EDs), beverages that contain high levels of caffeine in combination with other ingredients, have become alarmingly popular among children and adolescents around the globe in recent years. Given this rapid proliferation, it is critical to gain a more systematic understanding of aspects of ED use behaviors. Two large school-based datasets from two counties (U.S. and Germany) were used to addresses the following research questions: (1) to examine the cross-sectional and longitudinal associations between ED consumption and substance use (cigarettes, e-cigarettes, alcohol, and marijuana) among German and U.S. adolescents; (2) to examine the temporal associations self-reported ADHD symptoms and ED use among children and adolescents; (3) to evaluate the mediating role of ED use in the association between self-reported ADHD symptoms and substance use initiation. Study 1 found that current ED use was reported by roughly 20% of the participants in both adolescent samples. Those identified as Hispanic/Latino and multiethnic were more likely to use EDs, while those identified as Asian were less likely to use them. ED use was cross-sectionally and longitudinally associated with substance use outcomes (i.e., tobacco, hookah, alcohol, and marijuana). Study 2 demonstrated that ED use is cross-sectionally and prospectively associated with self-reported ADHD symptoms. Additionally, baseline ED use predicted more frequent ADHD symptoms at 12-months follow-up. Finally, Study 3 showed that ED consumption mediates the association between self-reported ADHD symptoms and substance use initiation. In sum, the findings of these studies contribute to a more comprehensive understanding of ED use behavior and its associated consequences. The research conducted in this dissertation has important implications for future research methods, theory, and ED prevention and intervention programs, while also may be beneficial in improving the lives of youth with ADHD symptoms.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Sociocultural stress, coping and substance use among Hispanic/Latino adolescents
PDF
Social self-control and adolescent substance use
PDF
Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults
PDF
The Internet activities, gratifications, and health consequences related to compulsive Internet use
PDF
Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California
PDF
Anxiety symptoms and nicotine use among adolescents and young adults
PDF
Affect, digital media use, physical activity, and ADHD in youth
PDF
Contextualizing social network influences on substance use among high risk adolescents
PDF
Exploring the role of peer influence, linguistic acculturation, and social networks in substance use
PDF
The role of social support in the relationship between adverse childhood experiences and addictive behaviors across adolescence and young adulthood
PDF
Smoke-free housing policies and secondhand smoke exposure in low income multiunit housing in Los Angeles County
PDF
Pathways of drug use among people who inject drugs
PDF
Factors and correlates of sexual behaviors among young adults from continuation high schools
PDF
Adolescent social networks, smoking, and loneliness
PDF
The dynamic relationship of emerging adulthood and substance use
PDF
Intrapersonal and environmental factors associated with Chinese youth alcohol use experimentation and binge drinking behaviors
PDF
The acute and longitudinal associations between sedentary behaviors, affective states, and emotional disorder symptoms among youth
PDF
An examination of the association between spousal support and type 2 diabetes self-management
PDF
Motivational interviewing with adolescent substance users: a closer look
PDF
Tobacco and marijuana surveillance using Twitter data
Asset Metadata
Creator
Galimov, Artur
(author)
Core Title
Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Preventive Medicine (Health Behavior Research)
Degree Conferral Date
2021-12
Publication Date
10/06/2021
Defense Date
07/08/2021
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
ADHD,Adolescents,Alcohol,energy drinks,gateway drug,OAI-PMH Harvest,substance use,tobacco
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Sussman, Steven (
committee chair
), Baezconde-Garbanati, Lourdes (
committee member
), Hanewinkel, Reiner (
committee member
), Huh, Jimi (
committee member
), Unger, Jennifer B. (
committee member
)
Creator Email
galimov@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC16022362
Unique identifier
UC16022362
Legacy Identifier
etd-GalimovArt-10140
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Galimov, Artur
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
ADHD
energy drinks
gateway drug
substance use