University of Southern California Dissertations and Theses

Intrapersonal and environmental factors associated with Chinese youth alcohol use experimentation and binge drinking behaviors

(USC Thesis Other)

Intrapersonal and environmental factors associated with Chinese youth alcohol use experimentation and binge drinking behaviors

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

INTRAPERSONAL AND ENVIRONMENTAL FACTORS ASSOCIATED WITH
CHINESE YOUTH ALCOHOL USE EXPERIMENTATION AND BINGE DRINKING
BEHAVIORS

by

Enrique Ortega

__________________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE)

August 2008

Copyright 2008 Enrique Ortega
ii
ACKNOWLEDGEMENTS
I would like to thank each member of my Dissertation Committee, Drs. C. Anderson
Johnson, Ping Sun, Lourdes Baezconde-Garbanati, Jennifer B. Unger, and Iris Chi for their
assistance and guidance in preparing this work of research. I am greatly indebted to Dr Ping
Sun for his thorough mentorship and great care he took in my development as a researcher.
The effort and accomplishment of this work is shared with Enrique, Rachel and Monet
Ortega, three incomparable engines for hope and goodwill that have never failed to propel
me forward. I am beholden to Marisol Romero for her immeasurable generosity and loving
companionship without which I would be a lesser person. I acknowledge each of my family
members and friends that have given me the love and support I needed to complete this
work.

iii
TABLE OF CONTENTS

Acknowledgements ii
List of Tables

iv
List of Figures

v

Abstract vi

Chapter 1: Introduction 1

Chapter 2: Background and Significance 9

Chapter 3: Study Methods 31

Chapter 4:

Cross-sectional Investigation of Intrapersonal and
Environmental Determinants of Alcohol Use in Chinese
Youth

41

Chapter 5:

Moderation Effects between Intrapersonal and
Environmental Determinants of Alcohol Use in Chinese
Youth

73

Chapter 6:

A Two Wave Follow up Latent Growth Curve Analysis of
Intrapersonal and Environmental Determinants of Alcohol
Use in Chinese Youth

97

Chapter 7: Conclusion 125
Bibliography

133

Appendix

Table A: Review of Empirical Studies of Adolescent
Alcohol use Determinants

144

iv
LIST OF TABLES
Table 1. Correlation Table of Variables of Interest.

47
Table 2. Summary of interested variables in the analysis among youth by gender

48
Table 3. Odds Ratio of Increasing Levels of AU by Predictors, Females

51
Table 4. Odds Ratio of Increasing Levels of AU by Predictors, Males

52
Table 5. Odds Ratio of Exclusive Levels of AU by Predictors of Interest,
Females

57
Table 6. Odds Ratio of Exclusive Levels of AU by Predictors of Interest, Males

58
Table 7. Relative Odds of Monthly Alcohol Use Among Female Lifetime Users

63
Table 8. Relative Odds of Monthly Alcohol Use Among Male Lifetime Users

64
Table 9. Relative Odds of Binge Drinking Among Female Monthly Users

65
Table 10. Relative Odds of Binge Drinking Among Male Monthly Alcohol Users

66
Table 11. Environmental and Intrapersonal Variable Interactions, Females

79
Table 12. Environmental and Intrapersonal Variable Interactions, Males

83
Table 13. Environmental and Intrapersonal Variable Interactions, Exclusive
Levels of Alcohol Use, Females

85
Table 14. Environmental and Intrapersonal Variable Interactions, Exclusive
Levels of Alcohol Use, Males

87
Table 15. Relative Odds of Monthly Alcohol Use among Female Lifetime
Alcohol Users, Interaction Analyses

89
Table 16. Relative Odds of Monthly Alcohol Use among Male Lifetime Alcohol
Users, Interaction Analyses

92
Table 17.

Relative Odds of Binge Drinking Among Female Monthly Alcohol
Users, Interaction Analyses
90
Table 18. Relative Odds of Binge Drinking Among Male Monthly Alcohol
Users, Interaction Analyses

93
Table A. Review of Empirical Studies of Adolescent Alcohol use Determinants 144

v
LIST OF FIGURES
Figure 1. Theoretical Model: Moderation Model for Youth Alcohol use
Consumption Behavior Change

8
Figure 2. Longitudinal Analysis of Stress, Monthly Drinking, 30 Day Smoking,
Females Only

104
Figure 3. Longitudinal Analysis of Stress, Monthly Drinking, Junk Food Use,
Females Only

105
Figure 4. Longitudinal Analysis of Stress, Monthly Drinking, Academic Score,
Females Only

107
Figure 5. Longitudinal Analysis of Hostility, Monthly Drinking, Academic
Score, Females Only

109
Figure 6. Longitudinal Analysis of Depression, Binge Drinking, 30 day
Smoking, Males Only

110
Figure 7. Longitudinal Analysis of Depression, Binge Drinking, Junk Food Use,
Males Only

112
Figure 8. Longitudinal Analysis of Depression, Binge Drinking, Academic
Score, Males Only

113
Figure 9. Longitudinal Analysis of Stress, Binge Drinking, 30 day Smoking,
Males Only

114
Figure 10. Longitudinal Analysis of Stress, Binge Drinking, Junk Food Use,
Males Only

115
Figure 11. Longitudinal Analysis of Stress, Binge Drinking, Academic Score,
Males Only

116
Figure 12. Longitudinal Analysis of Hostility, Binge Drinking, Academic Score,
Males Only
118

vi
ABSTRACT
This study investigated intrapersonal and environmental determinants of alcohol use in a
Chinese youth population to explore the role that such determinants have on alcohol use
behaviors. This study analyzed which determinants functioned as risk and protective factors
for the transition to increased alcohol use; and which variables moderated the impact of
other variables in their relationship to alcohol use onset and progression. Longitudinal data
from the China Seven Cities Study (CSCS), a health behavior study in seven of China’s
largest cities, were employed to study these matters. Data was analyzed from 14,434 7th,
8
th
, 10
th
, and 11th grade middle and high school students and their parents/guardians for this
study. This study identified important determinants of Chinese youth drinking onset and
transition behaviors. Junk food use, smoking behavior, and allowance were incrementally
associated with increasing levels of alcohol use in at least one of the analysis models in both
genders. Results also indicated that the associations between the selected intrapersonal
variables and both levels of alcohol use differed according to varying levels of academic
score, allowance, latchkey status, and parental education. In females results indicated that
monthly alcohol use reduced the experience of high stress at follow up when 30 day smoking
was analyzed in the model. In contrast results among males indicated that initial high levels
of stress were found to decrease binge drinking when junk food use and academic score was
analyzed in the same model. The results showed that while many alcohol use determinants
are similar to those found in the West, some may be particular to Chinese youth population.
The information presented here supports the view that alcohol use determinants must be
analyzed from as many adolescent life domains as possible given the fact that adolescent
alcohol use behaviors are affected by various life domains concurrently.
1
CHAPTER ONE: INTRODUCTION
Extensive research in the field of youth substance use (SU) has furthered our
understanding of the processes involved in alcohol use (AU) experimentation and the
development of alcohol use disorders (AUD). While several investigations have been
conducted to explain alcohol use behaviors in early adolescence (Jessor and Jessor 1977;
Elliott, Huizinga et al. 1985; Hawkins and Weis 1985; Simons, Conger et al. 1988; Kumpfer
and Turner 1990-1991; Petraitis, Flay et al. 1995) the majority of these investigations have
been carried out in the U.S. with primarily homogeneous study populations [Table A]. A
search of empirical studies seeking to identify risk and protective factors for alcohol use in
children and early adolescents revealed that 41 of 54 recent studies investigating alcohol use
behavior determinants in youth were conducted within the U.S. with an over representation
of Caucasian study populations [Table A].
Such investigations have been instrumental in identifying important risk and
protective factors for alcohol consumption behaviors in youth; nonetheless there remains a
great need to investigate the determinants of AU in populations where distinct socio-cultural
norms and regulations may have different influences on youth AU behaviors (Donovan,
Jessor et al. 1983; Donovan and Jessor 1985; Sher 2000; Donovan, et al. 2004; Sher, Grekin
et al. 2005). Investigations of this sort can help differentiate between the risk/protective
factors that are related to social and cultural norms and those related to other factors such as
biological and genetic predispositions to alcohol use (Lin and Cheng 2002). A number of
investigations have pointed out the importance of considering ethnic differences in alcohol
use investigations (Lin and Cheng 2002) (Hao, Chen et al. 2005). It is well established for
example that some Asian populations are less likely to become alcohol dependent compared

2
to other ethnic groups given particular physiological characteristics (Lin and Cheng 2002)
(Hao, Chen et al. 2005). It is estimated that approximately half of the Chinese population
exhibit a deficiency in aldehyde dehydrogenase (ALDH2), which is responsible for
metabolizing acetaldehyde into acetic acid, a component of the alcohol-metabolizing human
biochemical pathway (Lin and Cheng 2002). Polymorphisms of ALDH2 genes have been
found to modify alcohol-drinking behaviors and possible risks associated with alcoholism
(Lin and Cheng 2002).
China represents an ideal population to investigate youth alcohol use determinants
given these ethnic group considerations and the social and economic transition the country is
currently undergoing (Hao, Chen et al. 2005; Xing, Ji et al. 2006; Zhou, Su et al. 2006).
Investigators in China have stated that alcohol use among adolescents is a major problem
that is in strong need of studies that examine the patterns, trends, and consequences of use in
this population (Xing, Ji et al. 2006). Studies looking at youth alcohol consumption in China
have reported high alcohol use prevalence rates with over half of the populations sampled
indicating some alcohol consumption and equally alarming binge drinking behaviors (Hao,
Chen et al. 2005; Xing, Ji et al. 2006). Recent studies also indicate that adolescents in China
start using alcohol earlier than in most other countries with the exception of the United
Kingdom, Denmark, Finland, and the Russian Federation (Hao, Zhonghua et al. 2004).
These circumstances speak of the apparent need for alcohol use investigations that can
evaluate factors associated with alcohol use experimentation and with the transition of
alcohol consumption behaviors among Chinese adolescents.
Some of the more important adolescent AU determinants identified in the literature
pertain to both the intrapersonal and environmental domains of youths’ lives; these include:
parental monitoring of child behaviors, time spent in self-care, tolerant attitudes toward

3
deviance or approval of alcohol and drug use, affective disorders (depression, anxiety,
stress), engaging in other problem behaviors, and low bonding to conventional institutions
(Hawkins and Weis 1985; Hawkins, Catalano et al. 1992). These determinants are
commonly associated with alcohol use onset and experimentation but are seldom directly
tested for their involvement in adolescent alcohol use transition behaviors. Adolescent
substance use investigators have pointed to the fact that many etiological studies conducted
on adolescent SU have focused on psychosocial variables that predict initiation or that
distinguish users from non-users(Mayhew, Flay et al. 2000; Donovan et al. 2004; Glantz and
Mandel 2005; Sher, Grekin et al. 2005). These study outcomes may confound the process of
onset and escalation of use, and may fail to take into account that their might be different
predictors for SU onset and for SU transition behaviors (Maggs, Schulenberg et al. 1997).
Only 6 of a total 54 recent empirical studies identifying adolescent alcohol use determinants
analyzed longitudinal alcohol use transition behaviors in this population [Table A]. Of those
only 3 concentrated on alcohol use exclusively.
There are a host of issues particular to the problem of adolescent alcohol use that
may be beyond the scope of multi-substance use investigations that need to be addressed
with greater detail and focus (Donovan et al. 2004; Sher, Grekin et al. 2005). The
difficulties in analyzing common risk/protective factors associated with the use transition
behaviors of multiple substances are not only that dependence and outcomes are
characteristically different between substances, but that the risk/protective factors associated
with transition behaviors may be different for every substance. Thus we risk missing
important risk/protective factor nuances that may distinguish between use transition
behaviors among different substances. Reported risk factors associated with tobacco use
increase include: smoking dependence, intention to smoke, perceived stress, conduct
4
disorders, behavior problems, attachment to friends, and depression at different levels along
the smoking continuum (Glantz and Mandel 2005). Escalation of alcohol use has been
associated with parental alcohol involvement, parenting behavior, alcohol expectancies,
affiliation with alcohol-using peers, family dysfunction, stressful events and emotional
distress at different points along the use continuum (Power, Cynthia et al. 2005). Further,
we know that those who become dependent on illicit drugs differ from those who become
dependent on nicotine and alcohol in sex ratio, age, employment (Hughes 2006). There are
also marked differences between tobacco, alcohol and other drug use dependencies; tobacco
dependence for example almost never causes adverse behavioral outcomes such as violence,
traffic accidents, suicide, risky sexual behaviors, child neglect, etc (Hughes 2006) which
may also influence escalation of use between different substances.
This study proposed to investigate selected intrapersonal and environmental
determinants of youth alcohol use in a Chinese youth population to understand the role that
such determinants have on alcohol use experimentation and onset. This study also
investigated which of these determinants functioned as risk and protective factors for the
transition to increased alcohol use; and which variables moderated the impact of other
variables in their relationship to alcohol use onset and progression. This investigation’s
contribution to the field of youth alcohol use is that it will help us gain an increasingly
refined understanding of the etiological factors that contribute to AU onset and how these
factors interrelate with one another to contribute to progression along the gamut of
adolescent AU involvement. A more detailed understanding of the factors that influence
alcohol use initiation and consumption behavior change can help guide prevention
investigators in the field to focus on specific risk factors or combinations of risk and
protective factors which may promote delays of alcohol use onset or curb progression to
5
heavier use along different points of transition. Also of great importance is that this
investigation contributed to our knowledge of Chinese youth alcohol use behaviors; an area
of investigation that is in great need of advancement for the development of prevention
programs that take into account the determinants of Chinese youth alcohol use.
Longitudinal data from the China Seven Cities Study (CSCS), a health behavior
study in seven of Mainland China’s largest cities, were employed to study these matters.
Data was analyzed from 7th, 8
th
, 10
th
, and 11th grade middle and high school students and
their parents/guardians for this study.
Participant data was used to address the following specific aims:
(1) to identify intrapersonal and environmental predictors of alcohol use
experimentation, and of varying incremental levels of youth alcohol use
consumption behavior in a Chinese youth population.
(2) to investigate the potential moderating effects of intrapersonal dispositional
variables (depression, hostility, stress) on the relationships between environmental
variables (parental, economic, and school domains) and three level of alcohol use,
lifetime use, monthly use and binge drinking.
(3) to assess which intrapersonal and environmental domain variables contributed to
the progression of alcohol use consumption levels from a range of lifetime alcohol
use to monthly use and binge drinking behaviors employing 2 to3 waves of follow-
up data.
Theoretical Model
The multistage social learning model (MSLM) (Simons, Conger et al. 1988)
exemplifies an attempt to integrate various intrapersonal characteristics of adolescents,
parenting, and environmental factors into a single model of substance use. The MSLM
6
integrates intrapersonal characteristics, inadequate coping skills, and deficient interaction
skills into a comprehensive model that takes into account both distal and proximal causes of
general substance use (Simons, Conger et al. 1988; Petraitis, Flay et al. 1995).
The MSLM is a three stage model which describes the contributing factors of
adolescent substance use experimentation and transition to regular use and abuse. The first
stage hypothesizes that adolescent’s may become involved with substances if 1) they display
personal value systems that emphasize present-oriented goals over long term conventional
goals (e.g. education,); 2) have parents that provide infrequent supervision and discipline;
and 3) have parents who use substances. The second stage posits that adolescents will
become involved with substance using peers if they are deficient in social skills. Finally, the
third stage states that experimental substance use may escalate to regular and abusive use if
the adolescent displays emotional distress, has inadequate coping skills, observes their
parents engaging in substance use, and is involved with peers who encourage substance use
(Simons, Conger et al. 1988; Petraitis, Flay et al. 1995).
This study considered elements of the multistage social learning model (Simons,
Conger et al. 1988) to explicate adolescent alcohol use consumption patterns. This study
examined factors associated with transition use behaviors from alcohol use experimentation
to binge drinking behaviors. In this investigation I tested a framework of youth alcohol use
that incorporated the intrapersonal variables from the following domains: Psychosocial
(depression, stress, hostility), Lifestyle Habits (exercise, junk food use, appearance, health
status, smoking behavior), Leisure Time (TV time); and the environmental parenting
variables from the domains: Parenting (parent monitoring, latchkey, parenting style), School
(academic score, school confidence, school stress, study time use), Economic (allowance),
Parental (parent drinking, parent income, parent education), into a single analysis model.

7
This type of framework allowed me to tease out the pathways of greater influence
among the different variables that contributed to youth alcohol onset and possible
subsequent increased alcohol use consumption patterns. In addition this study identified any
moderation relationships that existed between the selected intrapersonal and environmental
variables. Figure 1 shows the moderation model that was tested in this study. In first
instance, this model tested the direct effect of the selected intrapersonal and environmental
variables on youth alcohol use while controlling for one another in the same model.
Secondly, this model explored the effects that intrapersonal dispositional variables had on
the relationship of environmental factors and alcohol use behaviors. While there is evidence
for the contributing effects of home and school environments on youth alcohol use an
important question to ask is how do intrapersonal dispositional variables affect this
relationship? There is a growing need to study the effects of biological and genetic
influences on all forms of health risk behavior. Indeed studies in several countries have
demonstrated that some behavioral characteristics present at childhood are associated with
alcohol-related problems later in adolescence. Investigations have shown that children are
more likely to develop alcohol use disorders later in life if they exhibit high levels of
novelty-seeking behavior, low levels of harm-avoidance behavior, are more aggressive, and
present early signs of depression (Rose 1998). While biological and genetic analyses were
beyond the scope of this paper one question this study posed is whether dispositional
possible proxy biological variables such as depression, stress and hostility, can change the
magnitude of association (moderates) of the relationship between environmental
determinants and alcohol use behaviors.

8
Figure 1. Moderation Model for Youth Alcohol use Consumption Behavior Change

Notes
1: Environmental variables:
Parenting: parent monitoring, latchkey, parenting style.
School: academic score, school confidence, school stress, study time use.
Economic: allowance.
Parental: parent drinking, parent income, parent education.
2: Intrapersonal variables:
Psychosocial: depression, stress, hostility.
Lifestyle Habits: exercise, junk food use, appearance, health status, smoking
behavior.
Leisure Time: TV time.
3: alcohol use consumption behavior change will is defined by looking at follow-up
associations of intrapersonal and environmental variables with three increasing levels of
alcohol use: experimental alcohol use, moderate use, and binge drinking

Environmental
Variables
1

Intrapersonal
Variables
2

Alcohol Use
Consumption
Behavior Change
3

9
CHAPTER TWO: BACKGROUND AND SIGNIFICANCE
Youth Alcohol Use Epidemiology, Burden of Disease
Despite norms and regulations prohibiting underage youth the use of alcohol,
adolescent alcohol use is widespread resulting in numerous negative health and social
consequences. As of 2005 lifetime alcohol use prevalence among 8
th
, 10
th
, and 12
th
grade
students in the US was reported to be 41%, 63.2% and 75.1% respectively (Johnston,
O'Malley et al. 2005). The rates among European children and adolescents are similarly
striking. The European School Survey Project on Alcohol and Other Drugs (ESPAD)
(Hebbell, Andersson et al. 2003) indicated that approximately 20% of 15 to 16 yr old
adolescents in Norway and Hungary and 50% in the UK, Czech Republic and Greece
reported 40 incidents of alcohol use or more during their lifetime.
The rates for alcohol intoxication are similarly high. A total of 19.95% of 8
th
grade
US students, 42.1% of 10
th
grade, and 57.5% of 12
th
grade students have reported being
drunk in their lifetime (Johnston, O'Malley et al. 2005). The proportion of European youth
who reported being drunk at age 13 or younger in 2003 ranged from 37% in Russia and the
UK to roughly 13% in Italy and Portugal (Institute of Alcohol Studies 2003). In Brazil a
study looking at secondary students across ten states reported that 30% of adolescents
between the ages of 10 and 18 yrs had used alcohol to the point of intoxication (Jernigan
2001). A total of 23% of Australian adolescents between 14 and 19 yrs reported consuming
seven or more standard drinks at least once a week in the 1998 National Drug Strategy
Household Survey (National Health and Medical Research Council 2001) (Jernigan 2001).
A national survey in Japan reported that 55% of adolescents between the ages of 13 and 17
yrs indicated drinking to the point of intoxication or unconsciousness (Jernigan 2001).

10
The potential harmful effects of youth alcohol use and misuse include additional
substance use experimentation and addiction, unintentional injuries, suicide, risky sexual
practices , violent assaults and victimization , traffic accidents, and alcohol poisoning among
others(Grant and Dawson 1997; Chen and Kandel 1998; Levy, Miller et al. 1999; Freisthler,
Gruenewald et al. 2003; Yoon, Yi et al. 2003). It is estimated in the US that nearly 20% of
traffic collisions involving drivers under the age of 21 involve alcohol use (Levy, Miller et
al. 1999). The total costs involved in underage drinking and traffic collisions per year total
over 19 billion dollars (Levy, Miller et al. 1999). According to the National Center for
Health Statistics (NCHS) 2% of alcohol poisoning deaths occurred in people younger than
21 years of age (Yoon, Yi et al. 2003). Of the approximately 2500 fatal underage suicide
victims in the US each year, it is estimated that 12% of male suicides and 8% of female
suicides are attributable to alcohol use (Levy, Miller et al. 1999).
Alcohol use in China
The most common forms of alcohol consumed in China include distilled liquor
(32%-54% ethanol), wine and yellow rice wine (12–18% ethanol) and beer (4–6% ethanol).
An additional form of alcohol consumed is medicinal liquor which includes traditional herbs
and is used to improve general health and treat a variety of ailments(Hao, Young et al. 1995;
Hao, Young et al. 1998a; Hao, Zhonghua et al. 2004; Hao, Chen et al. 2005)
Epidemiological surveys conducted in China have indicated that mean annual
alcohol consumption has increased steadily since 1980 and that such rates are approximating
those of western countries (Hao, Zhonghua et al. 2004). The increase in alcohol
consumption is believed to be a direct consequence of China’s rapidly developing economy
and increase in urbanization, where additional disposable income may facilitate social
drinking and the presence of alcohol in the home (Hao, Young et al. 1995; Hao, Young et al.
11
1998a; Cochrane, Chen et al. 2003; Hao, Zhonghua et al. 2004). A 2004 WHO sponsored
study of the general population (15 yrs and older) indicated that the proportion of past 3
month drinking rates were 63.8% for males, 18.3% for females and 43.8% for the total
sample. Drinking rates for the past year were reported to be 74.9% for males, 38.8% for
females and 59.0% for the total sample(W.H.O. 2004; Hao, Chen et al. 2005).
Total Chinese adult per capita consumption of pure alcohol in 1970–1972 was of
1.03 L; these rates rapidly increased to 5.17 L in just over two decades (1994–1996)(Xing, Ji
et al. 2006). The investigators pointed out that his percentage change in alcohol
consumption was second among 137 countries studied(Xing, Ji et al. 2006). The amount of
alcoholic beverages produced and sold in China has also increased steadily over the past
twenty years (Xing, Ji et al. 2006) (Hao, Young et al. 1995; Hao, Young et al. 1998a;
Cochrane, Chen et al. 2003; Hao, Zhonghua et al. 2004). China is the second largest beer-
producing country in the world (Hao, Young et al. 1995; Hao, Young et al. 1998a). In 2001
the sale of beer accounted for 73.1% of total beverage alcohol sales in China while spirits
accounted for 26% of total sales. As of 2001 more than four times the amount of ethanol was
consumed through spirits over beer (Hao, Chen et al. 2005). Taxation on imported liquors
(brandy and whisky) was cut in 1995 from 150% to 80% which made the volume of sales
double in the last decade(Hao, Chen et al. 2005).
Youth Alcohol Use in China
A current report in China has shown that alcohol use is common among middle
school students and that no reports are available which investigated the association between
alcohol misuse and behavioral risk factors (Xing, Ji et al. 2006). A national study of alcohol
use among 6th, 8th, and 10th grade adolescents reported that 69% of the total sample had
consumed one alcohol beverage during their life(Sen 2002). Xing (Hao, Chen et al. 2005)
12
conducted a survey among 115 middle schools, high schools and technical schools in
Shanghai and reported that 45.7% of the total population indicated drinking at least once in
their lifetime; the rates for past 30 day use were 17.8% and a total of 5.2 % reported being
intoxicated at least once in their life time(Hao, Chen et al. 2005). A survey conducted in
three middle schools in the Shijingshan district of Beijing reported that lifetime drinking
rates were 48.3% and 37.0% for male and female junior students respectively and 72.8% and
56.3% for male and female senior students. Surveys of the prevalence of alcohol use
conducted in Beijing, Wuhan, Hefei, and Shanghai provinces found that 40–53% of middle
school students reported using alcohol during their lifetime; 16–31% were current users, and
4–11% had experienced intoxication during the past year (Xing, Ji et al. 2006). Results from
the survey conducted in Beijing indicated that 36% of students in grades 7–11 were current
drinkers, 18% had engaged in binge drinking in the last month, and 38% initiated alcohol use
before 13 years of age(Xing, Ji et al. 2006).
In a study describing the frequency and patterns of alcohol use among 7-12 grade
students conducted in 18 provincial capitals (54,040) a total of 51.1% of students reported
ever using alcohol (male: 58.6%; female: 44.3%), 29.7% reported drinking before 13 years
of age, and 14.1% had been intoxicated at least once during the past year(Xing, Ji et al.
2006). In the past 30 days, 25.2% students reported consuming at least one drink of alcohol
(male: 31.4%; female: 19.6%) and 10.3% reported at least one episode of binge drinking
(male: 14.4%; female: 6.6%). Senior high school students reported a higher rate of binge
drinking (13.2%) and heavy binge drinking (2.2%) than junior high school students (binge
drinking: 5.8%; heavy binge drinking: 1.2%). The investigators reported reviewing a study
in China which showed that the percentage of those initiating alcohol use before age 13
years had been increasing from 1998 to 2003(Xing, Ji et al. 2006).

13
Need for Alcohol Use Research in China
While there have been abundant alcohol studies conducted in the West few studies
have been undertaken in Asian countries (Hao, Chen et al. 2005; Xing, Ji et al. 2006; Zhou,
Su et al. 2006) where drinking patterns and social/cultural norms are different (Janghorbani,
Ho et al. 2003). The findings reported in epidemiological studies in China suggest that
studies are needed to examine the patterns, trends, and negative consequences of alcohol use
among adolescents in that population. Research is needed that examines the predictive
utility of potential risk factors across cultures. This information could be used to adapt
preventive interventions that are tailored for adolescents of specific ethnic/racial, cultural, or
national backgrounds (Donovan et al. 2004). Even though a few number of risk factors have
been analyzed to understand the association between alcohol use behaviors in Chinese youth,
the influence of intrapersonal and environmental factors have rarely been investigated in this
population.
A number of Chinese researchers have noted that there is a great need to
systematically study and review alcohol related health behaviors in China (Hao, Chen et al.
2005; Xing, Ji et al. 2006; Zhou, Su et al. 2006). There have been few medical research
papers pointing out the disadvantages of alcohol drinking and propagating the harms of
alcohol abuse(Janghorbani, Ho et al. 2003). Few publications can be found on alcohol use
prevention programs (Hao, Chen et al. 2005) and few reports are available on the association
between binge drinking and other behavioral risk factors (Xing, Ji et al. 2006). Alcohol
related diseases are not major subjects for undergraduate and postgraduate education and
training in China and population based prevention has yet to be developed and routinely
implemented (Hao, Chen et al. 2005).

14
Youth Alcohol Use Risk Factors
Referring to the etiology of youth alcohol use there is extensive literature identifying
and associating selected intrapersonal and environmental risk factors to alcohol use
behaviors. In following the hypothesized determinants which have emerged from the field
of adolescent substance use behavior research, investigation efforts in underage alcohol use
have focused on the young individual’s intrapersonal characteristics, on their family life, on
their economic opportunities, and on their school experience (Hawkins, Catalano et al. 1992;
Zweig, Phillips et al. 2002).
The intrapersonal risk factors for youth alcohol use include depression (Deykin,
Levy et al. 1987) (Regier, Farmer et al. 1990; Grant and Harford 1995; Kumpulainen 2002)
(Gilman and Abraham 2001), stress (Abrams and Niaura 1987; Cooper, Russell et al. 1992;
Wills and Hirky 1996), and hostility (Windle and Windle 1995; Andersson, Magnusson et al.
1997). Additionally investigations have found that leisure time activities and lifestyle habits
can have important associations for youth alcohol use behaviors as well (Eccles and Barber
1999; Melnick, Miller et al. 2001; Rosal, Ockene et al. 2000). Risk factors from the
adolescent’s environment include parental characteristics such as how well they are
monitored, how much time they spend in self care and the level of control that their parents
have of their personal decisions. Other environmental risk factors include the academic
performance of the youth and their experience of school related stress; and general
demographic variables such as family economics and personal characteristics such as age,
gender and school grade enrollment.

15
Intrapersonal Risk Factors for Youth Alcohol Use
Depression
Results from the National Longitudinal Alcohol Epidemiologic Survey, a nationally
representative survey of 42, 862 individuals indicated that the lifetime odds of alcohol
dependence were 3.56 times higher for subjects presenting major depression according to
DSM-IV category than those with no major depression present (Grant and Harford 1995)
(Gilman and Abraham 2001). One study observed that individuals reporting past year
alcohol use report having severe depression symptoms at higher rates than the general
population (Grant and Harford 1995).
Likewise the risk of alcohol dependence in the general population has been noted to
be significantly higher among individuals with depression over those with no signs of the
disorder (Kessler, Nelson et al. 1996). Conversely, depression has been reported to be more
prevalent among individuals with alcohol dependence than without alcohol dependence
(Kessler, Crum et al. 1997).
Gilman(Gilman and Abraham 2001)reported that the odds of developing major depression at
one year follow up grew stronger with increasingly higher levels of alcoholic symptoms in
both females and males. This relationship was also observed for risk of alcohol dependence
which grew accordingly to higher levels of baseline depression symptoms. The association
between depression and alcohol use and misuse has been observed in children and
adolescents as well. Onset of substance use disorders tend to manifest around 15 years of age
if a mood disorder is present at earlier ages (Burke, Burke et al. 1994). A 3 year follow up
study analyzing depression and alcohol use in 12 year olds found depression to predict heavy
alcohol use at three years follow up from ages 7-9 (Kumpulainen 2000).

16
Stress
Intrapersonal etiological models of alcohol use commonly attribute the use of
alcohol as a relief of negative affective and emotional states (Sher, Grekin et al. 2005);
(Greeley and Oei 1999). One line of research has analyzed the use of alcohol as a method to
reduce the negative affective state of perceived stress (Cooper, Russell et al. 1992) (Fromme,
Stroot et al. 1993). Recent studies suggest that those individuals whose coping styles are
primarily limited to alcohol use as a means of combating stress comprise a potential
vulnerable population to alcohol use onset and misuse(Abrams and Niaura 1987; Wills and
Hirky 1996). Abrams & Niaura (Abrams and Niaura 1987) suggested that alcohol use may
develop in part due to the fact that an individual’s coping capacity or skills are not sufficient
to deal with situational demands. Consequently alcohol use outcome expectancies may
come into play in the desire to reduce stress or to delay or avoid stressful situational
demands (Abrams and Niaura 1987). Hussong, (Hussong 2003) reported males to be at a
greater risk for alcohol involvement associated with social adjustment and school problems
given a limited active coping style. In men heavy alcohol use was also associated to
relationship stress given a limited support seeking style.
Nonetheless, most studies of the relation between stress and alcohol use have
reported modest and inconsistent findings (Greeley and Oei 1999). Investigators have
pointed out that negative affect regulation from drinking is highly conditional upon both
intra-individual and situational factors (Greeley and Oei 1999; Sher, Grekin et al. 2005).
Hostility
A number of studies have shown that hostility and other problem behaviors in
adolescent populations are usually accompanied by substance use (Donovan and Jessor
1985; Griffin, Botvin et al. 2003). A higher level of alcohol consumption in youth has been
17
found to be associated with high levels of physical confrontations (Valois, Vincent et al.
1993); (Swaim, Deffenbacher et al. 2004)), higher levels of verbal aggression (Swaim,
Deffenbacher et al. 2004), and conduct disorder and norm breaking behavior (Boyle, Offord
et al. 1992). Weiner et al (Weiner, Pentz et al. 2001)found use of alcohol in the previous
month in grades 6 and 7 increased the odds eight fold at grades 11 and 12 of verbal and
physical hostile behaviors. Use of alcohol in the past month in grades 9 and 10 increased the
odds almost three times of hostile and disruptive behavior in grades 11 and 12.
Research has shown that early childhood aggressiveness and conduct disorder seem
to precede alcohol onset and alcohol use disorders (Clark, Pollock et al. 1997) (Jansen,
Fitzgerald et al. 1995). Investigators have reported aggressive behavior at time 1 to predict
alcohol use at time 2 in 12 year old adolescents (White, Brick et al. 1993). These findings
may be related to studies which have indicated that individuals with high externalizing
behavior traits present in childhood may be at risk for developing alcohol use disorders later
in life (Sher 2000).
Lifestyle Habits
Exercise
A clear and strong association has been established between sport, exercise, and
alcohol consumption (El-Sayed, Ali et al. 2005). Research has shown that participation in
sports has been associated with increased use of alcohol (Eccles and Barber 1999; Melnick,
Miller et al. 2001).
Alcohol is the most frequent substance used among athletes and habitual exercisers, and
alcohol related problems tend to appear more commonly among individuals with these
practices(El-Sayed, Ali et al. 2005).

18
It is believed that dopaminergic reinforcement mechanisms in the neural system that
are activated by substances such as alcohol are also activated during exercise(Thoren, Floras
et al. 1990). Behaviors such as drinking and exercise that increase the release of these
substances may lead to a positive response within the body and thus reinforce the behavior
associated with this positive response(El-Sayed, Ali et al. 2005). Consequently exercise may
produce similar effects to those experienced while consuming alcoholic beverages (Read and
Brown 2003). It has been suggested that exercise may facilitate better substance use
outcomes by providing a natural way for the alcohol user to achieve a pleasurable state that
does not require the use of exogenous substances(Read and Brown 2003).
A study that examined the associations between leisure-time exercise and a range of
health behaviors found leisure-time exercise was positively associated with daily alcohol use
among women only (Boutelle, Murray et al. 2000). The Disease Prevention and Health
Promotion Supplement to the 1985 U.S. National Health Interview (DUFOUR 1994) survey
found that moderate drinkers tended to live healthier lifestyles including engaging in more
strenuous exercise. An investigation conducted among a study population demographically
representative of the general U.S. population in which an empirical typology was created to
profile lifestyle patterns associated with non-clinical patterns of alcohol use found moderate
drinkers to consistently engage in healthy exercise patterns (SLATER, BASIL et al. 1999).
A brief, multi-health behavior intervention study integrating physical activity and alcohol
use prevention messages for high school-aged adolescents found demonstrated significant
positive effects at 3-months post-intervention for alcohol consumption, alcohol initiation
behaviors, use risk and protective factors, and exercise habits(Werch, Moore et al. 2005).

19
Junk food Use
An association has been established between higher levels of alcohol consumption
and poor dietary practices (Rosal, Ockene et al. 2000). A study that examined the co-
occurrence of risky health behaviors including poor diet and alcohol use among participants
enrolled in a randomized trial designed to test the effectiveness of a physician-delivered
counseling intervention on reduction in alcohol intake found that 28% of high risk drinkers
indicated having a poor diet (Rosal, Ockene et al. 2000). An investigation associating
lifestyle patterns with non-clinical patterns of alcohol use found fruit and vegetable
consumption among moderate drinkers to be second only to nondrinkers. In addition,
moderate drinkers were also second only to the nondrinkers in their ranking of healthy eating
as a personal value (Slater, Basil et al. 1999).
Appearance
Attitude toward appearance may be defined as the cumulative set of images,
fantasies, and meanings about the body, its parts and functions (Nieri, Kulis et al. 2005).
Attitudes toward appearance become increasingly important during adolescence as children
experience multiple physical and social changes (Nieri, Kulis et al. 2005). Research on girls
indicates that between 40 and 70% of adolescent girls are dissatisfied with 2 or more aspects
of their body (Neumark-Sztainer D 2002; Nieri, Kulis et al. 2005). A Minnesota study of
adolescents found that almost 26% of boys reported low body satisfaction (Guinn B 1997;
Nieri, Kulis et al. 2005).
Poor body image among adolescents has been associated with greater substance use.
Substance use may be a weight control strategy or a coping strategy for adolescents whose
severe body image problems are manifest (Nieri, Kulis et al. 2005). This concept is
supported by the findings of other studies linking poor body image to low self-esteem
20
and depression, which are risk factors for substance use (Nieri, Kulis et al. 2005).
Adolescents with poor body image may turn to substance use as an escape from their
feelings of low self-worth and depression. Substance use also may be a perceived avenue
toward social acceptance (Nieri, Kulis et al. 2005). Youth who dislike their looks or body
and assume that their peers feel the same about them may believe that using substances will
make them more attractive socially (Nieri, Kulis et al. 2005).
An increasing number of studies have pointed to image as an important factor in the
onset and maintenance of substance abuse among adolescents (Nieri, Kulis et al. 2005)
(Slater, Basil et al. 1999) An investigation associating lifestyle patterns with non-clinical
patterns of alcohol use found light drinkers to strongly disagree with the statement that
drinkers are attractive. This same study found moderate drinkers to somewhat disagree with
the statements that drinkers are attractive, and heavy drinkers to agree that drinkers were
attractive and more fun than non drinkers. Episodic drinkers were more likely to agree with
the statement that drinkers are more attractive or more fun than nondrinkers. Episodic
drinkers described themselves more strongly than any other drinking cluster as sexy, dating,
exciting, demanding, and clever (Slater, Basil et al. 1999). A study that explored body
image as measured by perceptions of weight and appearance and its impact on adolescent
drug use found boys who disliked their looks reported relatively greater amounts of recent
alcohol use (Nieri, Kulis et al. 2005). Anorexia nervosa and bulimia nervosa have been
associated with substantial rates of alcohol use disorder co-morbidity (Herzog 1999).
Health status
Early onset of alcohol use and heavy adolescent alcohol use has been associated
with both present and future negative physical and mental health consequences (Brook 2000;
Sabrina Oesterle 2004). Alcohol has a direct physiological effect due to its toxicity.
21
Long-term heavy use of alcohol has been seen to cause physical health problems in young
adulthood, especially among chronic heavy drinkers. Alcohol use also may affect health
indirectly as a result of its association with such behaviors as risk taking, delinquency and
violence, as well as other substance use, including smoking (Brook 2000; Sabrina Oesterle
2004).
A study examining the association of trajectories of heavy episodic drinking during
adolescence with health status and practices at age 24 found that young adults who did not
engage in heavy episodic drinking during adolescence had the lowest occurrence of health
problems and were most likely to engage in safe health behaviors at age 24 (Oesterle et al,
2004). Adolescent chronic heavy drinkers were more likely to be overweight or obese and
to have high blood pressure at age 24 than those who did not drink heavily in adolescence
(Oesterle et al, 2004).
A study conducted among a representative sample of 2235 adolescents’ ages 12–18
yrs conducted in Taiwan that examine the extent to which the use of alcohol, tobacco, and
betel nut related with health-related quality of life found that youth with recent alcohol use
tended to experience a poorer level of health-related quality of life. Alcohol use was also
shown to be associated with impaired levels of health-related quality of life in adolescents
(Chen and Storr 2006).
An investigation conducted among the general population living in five areas of
China in 2001 to health status related to drinking found that a significant difference existed
between the drinkers and the non-drinkers in terms of self-rated health status, 16.8% of the
drinkers and 22.6% of the non-drinkers considered that they were in very good health, and
3.0 % of the drinkers and 7.8 % of the non-drinkers thought that they were in bad or very
bad health (Hao, Zhonghua et al. 2004).

22
Smoking
Alcohol and tobacco are generally the first drugs with which adolescents
experiment, thus it is likely that initiation of drinking will be highly associated with smoking
(Jackson, Henriksen et al. 1997; Jackson, Sher et al. 2005). Alcohol use has been reported as
a reliable predictor of smoking; and tobacco use has been indicated as a marker for problem
drinking (Jackson, Henriksen et al. 1997; Jackson, Sher et al. 2005). Social drinkers and
alcoholics have been found to be more likely to smoke than nondrinkers (Jackson, Henriksen
et al. 1997; Jackson, Sher et al. 2005). Smokers have been found to be more likely than
nonsmokers to drink and to develop an alcohol use disorder (Jackson, Henriksen et al. 1997;
Jackson, Sher et al. 2005). Concurrent alcohol and tobacco use may interact to produce
health risks greater than the additive risks of each substance (Jackson, Henriksen et al. 1997;
Jackson, Sher et al. 2005).
Leisure Time
TV time
A higher level of television viewing has been associated with adolescent alcohol use
(Atkin 1990; Van Den Bulck and Beullens, 2005) Content analyses of alcohol use scenes on
television programs suggest that incidences of drinking by major characters occur
frequently, and that these portrayals generally present drinking as an activity that is
associated with happiness, social achievement, relaxation, and camaraderie (Thomsen and
Revke, 2006).
Furnham et al. (1997) (Furnham 1997) analyzed the portrayal of alcohol and
drinking in six British soap operas and found that 86% of all programs contained visual or
verbal references to alcoholic beverages; More alcohol was consumed than any other kind of
drink. An Australian content analysis of television programs found that 2.6 acts involving
23
alcohol were shown per hour in 1990 and 1997 (Atkin 1990; Van Den Bulck and Beullens,
2005) Everett et al. (1998) (Everett 1998)reported that 96% of US top grossing films had
positive references to alcohol consumption.
A study examining the influence of exposure to US-produced television programs on
the development of normative beliefs, expectancies, and intentions to drink among a group
of Norwegian non drinking adolescents reported that the influence of TV exposure was a
significant predictor of normative beliefs, expectancies and intentions to drink for those
subjects who reported having no friends who drink ( Thomsen and Revke, 2006). An
investigation that looked at television viewing and music video exposure predict alcohol
consumption among first and fourth year secondary school children of Flanders, Belgium
reported that overall television viewing per day and music television viewing at time 1
significantly predicted the amount of alcoholic beverages adolescents consumed while going
out at time 2 (Atkin 1990; Van Den Bulck and Beullens, 2005).
Environmental Risk Factors for Youth Alcohol Use
Home Domain
Poor parenting has been associated with alcohol use and misuse disorders in youth
populations including early onset of alcohol use, and higher levels of alcohol consumption
(Johnson and Pandina 1991), (Barnes 1984; Anderson and Henry 1994; Henry, Robinson et
al. 2003). The parenting practices that have been associated with contributing to alcohol use
disorders in youth are the level of monitoring the parents have on their offspring , the
amount of control they exert on their decisions, and the amount of time they leave their
children unsupervised.

24
Parental Monitoring
Parental monitoring refers to the amount of oversight parents have on their children.
Parental monitoring amounts to the level of supervision that parents have on their children’s
academic activities, the degree to which they know who their children associate with, and
the amount of knowledge that parents have on their children’s leisure time.
High parental monitoring has been reported to serve as a protective factor for
alcohol use onset and may prevent youth from associating with substance using peers
(Loukas and Prelow 2004) (Chilcoat, Dishion et al. 1995) (Dishion, Patterson et al. 1991;
Steinberg, Fletcher et al. 1994; Formoso, Gonzales et al. 2000; Hartos and Power 2000). In
a study looking at the association of parenting factors and adolescent problem behaviors
among 6
th
grade urban minority youth Griffin et al (Griffin, Scheier et al. 2000) found
parental monitoring to be associated with less alcohol consumption in males. Similar
findings have been reported among middle school urban youth with low levels of parental
monitoring predicting alcohol use experimentation (Getz and Bray 2005).
Despite cultural differences, similar findings have been reported for Asian parenting
practices. Li et al (Li, Fang et al. 2003) reported high parental monitoring among middle
school students in China to be associated with lower levels of alcohol use. Webb et al
(Webb, Bray et al. 2002)looked at gender, perceived parental monitoring, externalizing
behaviors, and adolescent alcohol use in a 2-wave longitudinal study among both middle and
high school students and found perceived monitoring to be related to less alcohol use over
time in both populations. Adolescent non-drinkers have been found to be more likely to have
parents who monitor how they spend their leisure time, check the time that they arrive home,
and to ask about adult supervision when they are not present (Beck, Shattuck et al. 1999).

25
Parenting Style
The concept of parenting styles refers to the level of involvement a parent has on the
child’s decisions and freedom of choice in family matters. The different styles of parental
involvement are authoritarian, permissive, and neglecting. Typical authoritarian parenting
styles involve the parents determining the child’s decisions with little or no input from them.
Permissive parenting styles are characterized by parent-child relationships where the child
acts under their own volition with minor parental control. Neglecting parenting styles entail
little or no care and involvement on behalf of the caretaker regarding their children’s
decisions (Kandel, Raveis et al. 1991) (Jackson, Henriksen et al. 1997).
A number of studies have noted that permissive parenting styles where the parents
do not set clear rules regarding substance use and behavior in general put children at risk for
early alcohol use experimentation (Jackson, Henriksen et al. 1997). A study looking at
young urban adolescents ages 9-17 found that parents who set clear limits on behavior and
exercised consistent authority had children with the lowest level of substance use (18%). A
total of 27% of youth experimented with substance use if they had parents categorized as
less permissive; 40% of youth experimented with substances if they had permissive parents
(Coombs and Landsverk 1988). A survey of eighth and ninth grade student’s in two public
school districts found tobacco and alcohol use to be associated with child perception of
lower parental authoritativeness; parent perception of parenting style was not associated
with child substance use(Cohen and Rice 1997).
Latchkey
Latchkey status or adolescent self-care refers to the time youth spend unsupervised
by their parents or adults, usually after school. Researchers have looked at the time a child
or adolescent spends unsupervised as a potential contributor to early alcohol use onset.
26
Higher amounts of time spent without parental supervision has been shown to be associated
with increased alcohol involvement among adolescents (Barnes, Reifman et al. 1994;
Steinberg, Fletcher et al. 1994; DiClemente, Wingood et al. 2001; Clark and Winters 2002;
LeDoux, Miller et al. 2002). An investigation looking at substance use among eighth grade
students who regularly care for themselves after school found that those who spent 5 to 10
h/wk in self-care were at almost twice the risk for alcohol consumption as those who did not
take care of themselves (Richardson, Dwyer et al. 1989). A study among middle school
students found that youth were 4 times more likely to consume alcohol if they spent 2 or
more days unsupervised by adults after school over those who were supervised 5 days per
week (Mulhall, Stone et al. 1996). Middle school and junior high school students who were
home alone two or more days per week were found to be more likely to have gotten drunk in
the past month than those youth who had parental supervision five or more times a week
(Mulhall, Stone et al. 1996).
School Domain
School Academic Performance
Investigation efforts have focused on the effects that adult expectations of high
academic achievement have on their children. Some investigations have shown that
increased substance use may occur among those with failing or substandard academic
attainments as an attempt to relieve depressive feelings (Kumpulainen 2002). General
substance use has been associated with lower academic achievement and negative
perceptions of school in as early as 6
th
grade US students (Sobeck, Abbey et al. 2000;
Ellickson, Tucker et al. 2001). These relationships have been observed among European
adolescents where academic underachievers were more likely to consume alcohol over those
with above average academic performance (Miller and Martin 1999).
27
A longitudinal study of adolescents (13 to 15 years) that investigated the
relationships between perceived academic performance and substance use found persistent
perceptions of academic failure at age 13 to predicted alcohol use at the two year follow up
point (Bergen, Martina et al. 2005). Increased alcohol use was also predicted among those
who developed a perception of academic failure despite an initial optimistic academic self-
assessment. Adolescents who perceived an academic improvement after a negative
assessment at baseline were at no increased risk for increased alcohol use at follow up
(Bergen, Martina et al. 2005). Similarly, a 3 year follow up study of 12 year old children
found perceived academic failure to predict heavy alcohol use at age 15 among girls
(Kumpulainen 2002).
The associations between alcohol use disorders and low academic achievement have
been reported to persist well into adulthood. A 31 year cohort study analyzed the
relationship between drunk driving offences and school performance/adult educational
achievements and reported that those who remained at the basic educational level to be at a
higher risk (3x for males, 7x for females) for being convicted of drunken driving in
adulthood (Riala, Isohaaai et al. 2003). Convicted drunk drivers were found to be more
likely to have lower marks in school and be left behind their age appropriate grades (Riala,
Isohaaai et al. 2003).
School Confidence, Study time
Educational level and school dropout have been found to be associated with the
development of alcohol abuse and dependence in adulthood (Crum, Ensminger et al. 1998).
Failure to meet expected levels of educational achievement may lead to problem behaviors
such as maladaptive drinking and alcohol dependence(Crum, Ensminger et al. 1998).

28
A study conducted to explore the characteristics associated with adult alcohol use
disorders among urban youth found that adolescents who work on their homework with a
family member weekly to every few months are significantly less likely to develop an
alcohol disorder in their early thirties relative to those who worked with a family member
less frequently. Failure to set rules about school behavior is also highly associated with risk
for alcoholism (Crum, Ensminger et al. 1998). This study found that predictors of
alcoholism included (1) those identified in first grade (the teachers' rating of
underachievement); (2) those identified in adolescence (few family rules about school,
infrequently working on homework with family, plans to limit schooling to high school); (3)
failure to complete high school (Crum, Ensminger et al. 1998).
School stress
A risk factor that is also of interest in the school domain is the role perceived school
related stress on youth alcohol use. Dissatisfaction with school experiences, ineffective
teaching, disruptive students, and parental expectation of academic performance may
contribute to the experience of stress and to ineffective coping mechanisms such as alcohol
use (Newcomb, Bemtler et al. 1986; Brook, Nomura et al. 1989; Bergen, Martina et al.
2005). A study looking at what factors predicted a variety of substance uses among a sample
of secondary students from two schools in Stirling, Scotland found that school stress was the
most important factor in predicting experimentation with alcohol use (Karatzias, Power et al.
2001).
Significance of the Proposed Study
While empirical studies associating risk and protective factors to adolescent drinking
behavior relate important findings regarding determinants of use and misuse, many fail to

29
distinguish between risk factors for initiation and risk factors for transition behaviors along
the dimension of alcohol involvement in adolescence (Donovan et al. 2004).
An investigation for youth drinking behaviors is necessary which can allow
researchers in the field to identify factors that are associated with both alcohol use
experimentation and with progressive consumption behaviors.
A review of pertinent literature indicated that although a number of empirical
investigations have been conducted to explain factors influencing adolescent substance use
(Jessor and Jessor 1977; Hawkins and Weis 1985) (Simons, Conger et al. 1988; Kumpfer
and Turner 1990-1991; Petraitis, Flay et al. 1995) few have tested these constructs on
adolescent alcohol use behaviors exclusively (Donovan and . : 2004). Understanding how
particular risk factors for underage alcohol use interrelate with one another and influence the
alcohol use behaviors of adolescents can serve investigators greatly when considering
methods and points of intervention for youth alcohol use prevention efforts.
In this study I analyzed the interrelationships among different youth psychosocial
variables and environmental influences that had the possibility to influence the onset of
alcohol use and to increase consumption patterns in a Chinese youth population. This
allowed me to investigate which psychosocial states shared common causes with Chinese
adolescent alcohol problems and to identify which factors moderated the relation between
certain variables and the development of alcohol use and abuse in youth.
Research in the field has shown that alcohol use experimentation commences
between the ages of 13 and 15 years and in some cases even earlier (Lewinsohn, Rohde et al.
1999; Degenhardt, Hall et al. 2000). As reported here we know that adolescents in China
start using alcohol earlier than in most other countries (Hao, Zhonghua et al. 2004). If we

30
couple this information with the fact that alcohol use disorders have been reported to have a
modal onset between the ages of 15 and 19 yrs (Moss and Lynch 2001) we can see the
critical significance of investigating the factors and conditions that lead to alcohol
experimentation and increased consumption behavior change in early ages. Effective youth
alcohol use prevention efforts must consider the potential contributors to this behavior from
the most salient domains of an adolescent’s life. The knowledge of the role that potential co-
morbid psychosocial disorders may have on alcohol use behaviors can help direct
investigators to focus intervention efforts on the underlying influences of alcohol use and on
concurrent psychosocial conditions.

31
CHAPTER THREE: STUDY METHODS
Data source and sample selection
The data used in this study are from a health behavior study in seven of Mainland
China’s largest cities (China Seven Cities Study). The primary purpose of the China Seven
Cities Study (CSCS) is to assess the specific influences on tobacco use in China that will
lead to the development and assessment of community-based approaches to tobacco use
prevention and control. The secondary purpose is to better understand the role of rapid
social, economic, and cultural change on tobacco use and related health practices and
outcomes. The choice of cities for this study, from the far northeast to the far southwest of
China’s population centers, permits the assess of relationship of economic and social factors
and their influence on smoking and health outcomes. These seven cities are located in four
regions in China: North-East (Harbin, Shenyang), Center (Wuhan), South-West (Chengdu,
Kunming), and Coastal (Hangzhou, Qingdao).
Procedure
The baseline survey of the CSCS was conducted between October 2002 and
December 2002. In six of the seven participating cities, the municipal Center for Disease
Control and Prevention (CDCP) is the major partner of the CSCS. In Kunming, the Health
Education Institute (HEI) is the major partner, The director and staff of the CDCP (or HEI
in Kunming), leadership from the municipal Health Bureau, and members of the Education
Committee provided assistance to the U.S. research team in gaining access to the study
schools, providing input during questionnaire development, obtaining informed parental
consent, delivering a written and verbal youth script (in lieu of a written assent form), and
providing data collection teams. The CSCS includes surveys of three age groups: middle
and high school students, college students, and parents of middle and high school students.

32
School selection
Prior to the survey for the middle and high school students, students took home
consent forms and questionnaires for their parents to complete. Students returned the
consent forms and parents questionnaires to their classroom teachers in sealed envelopes.
Because individual youth-student assent documentation is not widely used in China and may
be disruptive to cultural beliefs, we utilized an oral assent process with a standardized script
before the classroom survey for the middle and high school students. The institutional
review boards established in each of the seven participating Chinese cities and the U.S.
Office for Human Research Protections (OHRP), with whom we consulted, approved of this
method. The script, which appeared on the cover page of the youth version of the
questionnaire, was read aloud verbatim to each participating class of middle, and high
school, and professional/vocational students before questionnaire administration by either a
data collector or other trained personnel from each city’s CDCP of HEI. The script
explained the study objectives and procedures before the survey. Students who decided not
to participate in the survey were dismissed by the data collectors even if their parents had
signed parental permission forms. The students completed the surveys in their classroom
sessions proctored by the data collector, without the schoolteachers present. For college
students, surveys were also conducted in the classroom and consent was obtained prior to
survey administration. Participants were asked not to write any information that might
identify them on the questionnaires. All study procedures and survey instruments were
approved by both the University Southern California and Chinese Institutional Review
Boards.

33
Participants
This analysis only involves data collected from the Youth and Adults surveys, thus
only participants to these two surveys are described here. Stratified sampling strategy was
adopted for the baseline survey to select middle and high school samples. In each city, three
administrative districts with the highest, lowest, and medium average income of the residents
were to be selected. Local Education Committees provided a list of middle and high schools
grouped according to three levels of academic achievement within each district. One middle
school and one traditional high school were randomly selected from each of the 9 (3 levels ×
3 districts) clusters to participate in the study. One class from each of grades 7 and 8
(middle school) or 10 and 11 (high school) were recruited for the study. For the professional
high schools, only one school was selected from each district. These professional schools
were matched across districts on number of enrollments, type of occupational training, and
male/female student ratios. Two academic majors were randomly selected and students in
each of the 10
th
and 11
th
grades in these majors were recruited. Across the seven cities, a
total of 15,516 7th, 8
th
, 10
th
, and 11th grade middle and high school students and 31,032
parents/guardians of these invited students were invited to take part in the study. A total of
14,434 (93.0%) middle and high school students completed the survey. A total of 25,300
parents/guardians (12,445 males and 12,855 females) of the students completed the adult
survey. This analysis included non-missing data from all students in the Youth survey and
the Adult survey. Among the 14,434 students who took part in the student survey, 11052
students were retained for this analysis.

34
Survey instruments
Two versions of the self-administered paper-and-pencil questionnaire were
developed for the survey: Youth (for 7
th
and 8
th
middle schools; 10
th
and 11
th
grade high
schools, and 10
th
and 11
th
grade professional/vocational students), and Adults (for
parents/guardians of students participate in the Youth survey). Survey items were translated
from English to Mandarin by translators fluent in both languages and trained in behavioral
science theory and research. The final versions of the questionnaires were approved by the
Education Committees and CDCP - HEI research teams in each of the seven participating
Chinese cities prior to administration.
The questionnaires incorporated items from the U.S. Center for Disease Control’s
Behavioral Risk Factor Surveillance System (BRFSS - adults), Youth Risk Behavioral
System (YRBS - youth), 1995 Youth Risk Behavioral System (YRBS – college),
Transdisciplinary Tobacco Use Research Center (TTURC) Wuhan Smoking Trial survey
(Unger, Chen et al. 2001), and other sources and addressed topics of tobacco use, alcohol
and drug use, STD/HIV risks, relevant dispositional factors (depression, stress, hostility),
health communication channels, public awareness of tobacco policies,
individualism/collectivism, western influences, family structure, health status, and
developmental and demographic characteristics.
Measures
Parental monitoring. The students responded to one item about how closely their parents
monitor their activities (D. A. Cohen, Richardson, & LaBree, 1994), “How often do your
parents check whether you’ve done your homework?” The scale for this item included: “very
often”, “sometimes”, “hardly ever”, and “never.” The scale was inverted such that a higher
score would indicate closer parental monitoring.

35
Latchkey status. This variable was assessed by the following question: “How many days in a
week do you find that there's no adult at home when you go home after school?”. Response
choices ranged from: 1=“Never”, 2=”1-2 days a week”, 3=”3-4 days a week”, and 4=”5 days
a week”.
Parenting Style. The students responded to a single item, “In general, when it comes to
spending money, doing fun activities, and deciding how late you can stay out, which of these
statements most closely describes how you and your parents make decisions?” The scale for
this item included: “My parents generally make these decisions”, “My parents ask for my
opinion but they generally make the decisions”, “I ask for my parents' opinion, but I
generally make the decisions”, “I generally make the decisions” (Radziszewska, Richardson,
Dent, & Flay, 1996). This scale was inverted such that a higher score indicated more
involvement by the parent in the student’s activities.
Depression. Three questions adapted from the Center for Epidemiological Studies
Depression Scale, CES-D (Radloff 1977). CES-D is a 20-item self-report measure that uses
4-point scales to tap depressed mood over the past week. To assess the depressive
symptoms, students and their parents in this study were asked: "Think about how you felt
during the past 7 days. On how many of these days did you have trouble shaking off sad
feelings?", "On how many of these days did you feel depressed?", and "On how many of
these days did you feel sad?" Response options ranged from: 1="0-1 day", 2= "2-3 days",
3= "4-5 days", and 4= "6-7 days". The final score was the average of the four responses,
with a possible range of 1 to 4. Alpha for this depression score was 0.84 in the Youth survey
and 0.86 in the Adult survey.
Stress. The perceived stress measure was derived from Cohen, Kamarck and Mermelstein's
10-item scale (Burt, Cohen et al. 1988); (Cohen, Kamarck et al. 1983). Five items of the 10
36
items were pilot tested from the scale, three items with the highest item-total correlations
were selected for the survey. The three items were: In the last month, how often have you 1)
felt nervous and "stressed", 2) that you could not cope with all the things that you had to do,
and 3) felt difficulties were piling up so high that you could not overcome them. The
possible answers to these three questions were 1. Never (0 days), 2. Almost never (1-2
days), 3. Sometimes (3-5 days), 4. Fairly often (6-15 days), and 5. Very often (15+ days).
The final score for the perceived stress was the average over these three answers, with a
possible range from 1 to 5. Alpha for this perceived stress was 0.83 in the Youth survey and
0.84 in the Adults survey.
Hostility. To assess self-reported hostility, the following 3 questions adapted from the Buss-
Durkee Hostility Inventory (Buss and Durkee 1957) were asked: "I lose my temper easily",
"I can't help being a little rude to people I don't like", and "Lately, I have been kind of
grouchy". Responses were rated on a 4-point scale: 1= "definitely no," to 4= "definitely
yes". The final score was the average of the three responses, with a possible range of 1 to 4.
Cronbach's alpha for this hostility scale was 0.67 in the YOUTH survey and 0.75 in the
ADULTS survey.
School stress. School stress was assessed with the following two items: “How often do you
feel overwhelmed by your schoolwork?” Response choices ranged from “Very Often” to
“Never”. Next, “How confident are you that you will achieve your academic goals?”.
Response choices ranged from “Extremely confident” to “Extremely unconfident”.
Academic performance. Academic performance was assessed by asking the students: “What
is your usual academic performance at your current school or the last school where you
received grades?” The following response choices were provided: “Mostly A’s, or 90 or
more points, or Superior”, “Mostly B’s, or 80-89 points, or Very Good”, “Mostly C’s, or
37
70-79 points, or Average”, “Mostly D’s, or 60-69 points, or Below Average”, “Mostly F’s,
or Below 60 points, or Failing”. This scale was inverted such that a high score in the scale
indicated lower academic scores.
Exercise. Exercise was assessed by asking participants “How many times a week do you
breathe hard and sweat for over 20 minutes while riding a bicycle, walking fast, jogging,
dancing, or doing other exercise or hard physical labor?” Response choices ranged from
“None” to “8 or more times”
Junk food use. A scale was created to create this variable. This variable was created by
frequency of consumption of sweets, frequency of consumption of snacks, and frequency of
consumption of fast food. The final score was the average of the three responses, with a
possible range of 1 to 6.
Appearance. Attitudes toward appearance was assessed by asking the participant to provide
answers for the following three questions: “It's important for people to work hard on their
figures/physiques if they want to succeed in current society”; “Attractiveness is very
important if you want to get ahead in our society”; “In current society, it is important to
always look attractive”; Response choices ranged from “Strongly disagree” to “Strongly
Agree” The final score was the average of the three responses, with a range of 1 to 5.
Health Status. Health status was created by asking participants to answer the following
question: “How would you describe your health?” Response choices ranged from “1.
excellent” to “4. poor.” This item was inverted such that a higher score would indicate
better indices of health.
Lifetime smoking. Students responded yes or no to the following item, “Have you ever tried
cigarette smoking, even a few puffs?” Lifetime smoking measured at baseline was included
in the regression as a dichotomous control variable.
38
Past 30 day smoking. Students responded to the following item, “During the past 30 days, on
the days you smoked, how many cigarettes did you smoke per day? The response options
included, “I did not smoke cigarettes during the past 30 days”, “Less than 1 cigarette per
day”, “1 cigarette per day”, “2 to 5 cigarettes per day”, “6 to 10 cigarettes per day”, “11 to
20 cigarettes per day”, and “More than 20 cigarettes per day.” The baseline and follow-up
data were dichotomized to account for highly skewed data. A student was categorized as a
past 30-day smoker unless he or she marked only the first response item. The baseline data
for past 30 day smoking was included in the regression model as a control variable, and the
one-year, follow-up data was included as the outcome variable.
TV time. This variable was assessed by asking “On school days, how much time on
average do you spend watching TV/video?” Response choices ranged from “1. None/Hardly
any” to “7. more than 4 hours”
School Confidence.
Study time use. Study time use was assessed by asking the question: “On school days, how
much time on average do you spend doing homework after school?” Response choices
ranged from “1. None/Hardly any” to “7. more than 5 hours.”
Allowance. Allowance was assessed by enquiring: “How much allowance per week do you
have the freedom to spend anyway you want?” Response choices ranged from: “1. None” to
10. More than 90 Yuan.”
Experimental alcohol use. Participants were asked: During your life, on how many days
have you had at least one drink of alcohol? Response choices included incremental range of
days from “0” to “100 or more days.”
39
Moderate alcohol use. Participants were asked: During the past 30 days, on how many days
did you have at least one drink of alcohol? Response choices included incremental range of
days from “0” to “30 or more days.”
Binge drinking. Participants were asked to indicate the number of days during the past 30
days, that they had have 5 or more drinks of alcohol in a row within a couple of hours?
Response choices included incremental range of days from “0” to “20 or more days.”
Other variables assessed and used as covariates in this analysis included grade, city,
type and academic level of school attended (poor, medium, good, professional/vocational),
economic level of municipal district (poor, medium, good), parental education level,
parental income level, and parental binge drinking.
Parental education level was obtained by asking both parents to indicate the last
degree of education completed. Response choices included: “1. Didn't go to elementary
school or didn’t graduate from elementary school”; “2. Elementary school graduate”; ”3.
Junior high school graduate”; “4. Senior high school graduate”; “5. Training/vocational
school graduate”; “6. College graduate”; “7. University graduate or higher”. The highest
education level of one parent was used for this variable.
Parental income level was the average income reported from both parents. We
assessed parental income with the question “What is your total monthly family income from
all sources?” A total of 11 income categories were provided which ranged from “1.<100
yuan” to “11. more than 10,000 yuan”.
Finally, the maximum level of heavy drinking by either parent was used to indicate
parental binge drinking per family. Each parent was asked to consider all types of alcoholic
beverages they drank in the past 30 days and answer how many times during the past 30
days did they have 5 or more drinks on an occasion? Response choices were: “1. none”; “2.
40
1 time”; “3. 2 times”; “4. 3 to 5 times”; “5. 6 to 9 times”; “6. 10 to 19 times”; “7. 20 or more
times.”
Data Analysis
Analyses for studies one and study two were performed on cross-sectional data from
the CSCS baseline survey. Analyses for study three were performed on cross-sectional data
from CSCS baseline survey and data from the CSCS survey wave 2. Analyses were
performed with SAS (v 9.1; SAS Institute, Cary, NC) and Mplus (v5 Muthen & Muthen).
Univariate characteristics of all study variables were reported by descriptive statistics
including means, standard deviations and percentages. Gender differences for each study
variable were tested by t-test for continuous variables, and χ
2
test for categorical variables.
Power Analysis
Due to the large sample size available for this study, small effects can be statistically
significantly detected for hypotheses testing. In our analysis, the independent variables (e.g.
dispositional variables) were compounded with measurement errors, and the dichotomous
assessment of alcohol uses (Yes/No) will be used as the outcomes. A measurement error
considered power analysis for multiple logistic regressions (Tosteson, Buzas et al. 2003) was
considered appropriate for the power calculation for this study. Assuming two sided
alpha=0.05, power=0.80, and 0.50 as the reliability of measures in the independent variables
(e.g. depression), it was calculated that the available sample size of 14,434 youth would be
enough to detect an Odds Ratio as low as 1.11 for each 1 standard deviation change in the
independent variables (1sdciid). Under the same assumption, analyses conducted by gender
(males or females exclusively) would be able to detect an Odds Ratio as low as 1.14 for
1sdciid. Similarly, for within gender interaction analysis, the sample size was enough to
detect an OR of 1.31 1sdciid for difference in effects between evenly divided sub-groups.

41
CHAPTER FOUR: CROSS-SECTIONAL INVESTIGATION OF INTRAPERSONAL
AND ENVIRONMENTAL DETERMINANTS OF ALCOHOL USE IN CHINESE
YOUTH
A fair number of empirical studies that have sought to identify risk and protective
factors for adolescent drinking have reported important associations between intrapersonal
and environmental determinants and youth alcohol use. However, investigators have stated
that many studies that analyze adolescent alcohol use determinants continually fall short in
distinguishing between risk factors for initiation and risk factors for transitions along the
dimension of alcohol involvement (Donovan and . : 2004). Another shortcoming of current
youth alcohol use investigations is that the majority of investigations that have been
conducted to explain alcohol use behaviors in early adolescence (Jessor and Jessor 1977;
Elliott, Huizinga et al. 1985; Hawkins and Weis 1985; Simons, Conger et al. 1988; Kumpfer
and Turner 1990-1991; Petraitis, Flay et al. 1995) have been carried out in the U.S. with
primarily homogeneous study populations [Table A].
There is a great need to investigate adolescent alcohol use determinants in
populations where distinct socio-cultural norms and regulations may have different
influences on youth alcohol use behaviors (Donovan, Jessor et al. 1983; Donovan and Jessor
1985; Sher 2000; Donovan et al. 2004; Sher, Grekin et al. 2005). Investigations that focus
on diverse socio-cultural norms and regulations can help differentiate between the risk and
protective factors that are related to social and cultural practices and those that are related to
other factors such as biological predispositions.
Investigators in China have stated that alcohol use among adolescents is a major
health problem. The problem of youth alcohol use in China is in great need of investigations
that can evaluate factors associated with alcohol use experimentation, and with those related
42
to the transition of alcohol consumption behaviors. Studies are needed which examine the
patterns and trends of alcohol use behaviors in this population (Xing, Ji et al. 2006).
A review of adolescent substance use literature has indicated that although a number
of empirical investigations have been conducted to explain factors influencing adolescent
substance use (Jessor and Jessor 1977; Hawkins and Weis 1985) (Simons, Conger et al.
1988; Kumpfer and Turner 1990-1991; Petraitis, Flay et al. 1995) few have tested these
constructs on adolescent alcohol use behaviors exclusively (Donovan and . : 2004).
Understanding how particular risk factors for youth alcohol use interrelate with one another
and influence alcohol use behaviors can prove critical when considering methods and points
of intervention for youth alcohol use prevention efforts.
An investigation of youth drinking behaviors is necessary which will allow
researchers in the field to identify factors that are associated with both alcohol use
experimentation and with progressive consumption behaviors. As a first step this study
investigated selected intrapersonal and environmental determinants of youth alcohol use in a
Chinese youth population to understand the role that such determinants had on alcohol use
behaviors including experimentation and binge drinking behaviors.
In this first study I analyzed the associations of different youth psychosocial and
environmental domain factors with the alcohol consumption behaviors of a Chinese youth
population. This study tested the direct effect of the selected intrapersonal and
environmental variables on youth alcohol use while controlling for one another in the same
model. It is important to note that the majority of empirical studies investigating the
influence of intrapersonal and environmental determinants of alcohol use do so while
concentrating on one domain exclusively. I believe that it is of critical importance to study
both intrapersonal and environmental domain factors in the same theoretical model in order
43
to better understand the influence that each of these variables has on one another in their
relation to alcohol use consumption behaviors. While investigations have shown the
influence of parental practices on youth consumption behaviors, knowing what effect a
concurrent psychosocial variable may have on such behaviors is important to understand.
One way to answer this question is to analyze the effects of both variables in the same
analysis model. This will give us a more complete understanding of how concurrent daily
life factors affect adolescent AU behaviors.
This study considered the principal intrapersonal and environmental factors that
have been implicated in Western youth alcohol use behaviors for analysis. This study
examined the cross-sectional associations between the intrapersonal domain variables:
Psychosocial (depression, stress, hostility), Lifestyle Habits (exercise, junk food use,
appearance, health status, smoking behavior), Leisure Time (TV time); the environmental
domain variables: Parenting (parent monitoring, latchkey, parenting style), School
(academic score, school confidence, school stress, study time use), Economic (allowance),
Parental (parent drinking, parent income, parent education), and 3 distinct levels of alcohol
use in youth. The three levels alcohol use analyzed were: lifetime use, monthly use, and
binge drinking. The variables from all three domains were tested in a single model in order
to determine their contribution to each level of alcohol use respectively, while controlling for
the effects of one another.

44
Hypotheses
Consistent with the reviewed theories and empirical studies that have implicated the role of
intrapersonal, family, and school factors in youth alcohol use in mostly Western populations,
the following findings were expected to be observed in our sample of Chinese youth:
(1) Higher levels of depression and hostility would be associated with increasing
levels of alcohol use.
(2) Higher levels of exercise and junk food use would be associated with increasing
levels of alcohol use.
(3) A greater concern for appearance and poor health status would be associated
with increasing levels of alcohol use.
(4) Smoking behavior would be associated with increasing levels of alcohol use.
(5) Higher levels of parental monitoring and authoritarian parenting styles would be
associated with lower levels of alcohol use.
(6) Higher amounts of time spent in self-care (latchkey status) would be associated
with increasing levels of alcohol use.
(7) Poor academic performance and high school stress would be associated with
increasing levels of alcohol use.
(8) Greater levels of school confidence and more time spent on school work would
be associated with lower levels of alcohol use.
(9) Greater disposable income would be related to increasing levels of alcohol use.
Analysis
Analyses for this study were performed on cross-sectional data from China Seven
Cities Study (CSCS) baseline survey. Due to the clustering of students within schools and
the possible intra-school correlation between students within the same school, a generalized
45
linear model was applied in the analysis. In this model, both inter-school and intra-school
correlations were assumed for self-reported measure, and a logit link function was assumed
to transform the dichotomous outcome.
I analyzed the cross-sectional associations between the intrapersonal risk factors:
psychosocial (depression, stress, hostility), lifestyle habits (exercise, junk food intake,
appearance, health status), leisure time (TV time); the environmental risk factors: home
domain (parental monitoring, parenting Style, latchkey), school domain (school academic
performance, school stress, school confidence, study time), economic domain (allowance),
parental domain (income, education, alcohol use) and three levels of alcohol use (lifetime,
monthly, and binge drinking).
All independent variables were tested in a single model in order to control for
individual effects. Possible confounders adjusted for in the regressions included age, grade,
puberty, city, type and academic level of school attended (poor, medium, good,
professional/vocational), economic level of municipal district (poor, medium, good),
parental education level, parental income level, and parental heavy alcohol use. Analyses
were conducted separately by gender in order to assess the gender differences that existed in
these associations.
Three models were tested in order to fully understand the relationships that existed
between the predictors of interest and each level of alcohol use. In the first model (general
alcohol use model) 3 dichotomous levels of alcohol were tested: lifetime alcohol use
(yes/no), monthly alcohol use (yes/no), and binge drinking (yes/no). Statistical associations
between each intrapersonal and environmental variable and each level of alcohol use were
analyzed. This model evaluated the relative odds of the variables of interest predicting a
level of alcohol use between those who endorsed a level of use and those who did not.
46
A second model was tested (exclusive alcohol use model) in which exclusive
alcohol use categories were used. Three levels of use were used which were exclusive of one
another. The lifetime use category for example only comprehended those users who
endorsed lifetime use but not monthly use nor binge drinking. This model evaluated the
relative odds of our variables of interest predicting a level of alcohol use between those who
endorsed a single level of use and non alcohol users.
Lastly a third model was tested (incremental alcohol use model) where I analyzed
which variables predicted transition from one level of alcohol use to an incremental level of
use (e.g. life use to monthly use or monthly use to binge drinking). Two models were tested,
one in which the relative odds of monthly alcohol use among lifetime alcohol users was
assessed, and another in which the relative odds of binge drinking among monthly alcohol
users was assessed.

Table 1. Correlation Table of Variables of Interest.

Depression Stress Hostility Lifetime AU Monthly AU Binge drinking Monitoring Latchkey Parent Ed level P. income P. Binge Academic score
Depression 1
Stress 0.53 1
<.0001
Hostility 0.29 0.31 1
<.0001 <.0001
Lifetime AU 0.11 0.12 0.06 1
<.0001 <.0001 <.0001
Monthly AU 0.13 0.07 0.06 0.47 1
<.0001 <.0001 <.0001 <.0001
Binge drinking 0.12 0.04 0.05 0.32 0.6 1
<.0001 <.0001 <.0001 <.0001 <.0001
Parent monitoring -0.08 -0.15 0.004 -0.16 -0.08 -0.07 1
<.0001 <.0001 0.56 <.0001 <.0001 <.0001
Latchkey 0.08 0.07 0.03 0.05 0.05 0.04 -0.11 1
<.0001 <.0001 0.0004 <.0001 <.0001 <.0001 <.0001
Parent education level 0.003 0.04 -0.03 -0.004 -0.04 -0.05 -0.02 0.04 1
0.726 <.0001 0.003 0.5803 <.0001 <.0001 0.001 <.0001
Parental income -0.01 0.03 -0.008 0.02 -0.003 -0.003 -0.08 0.07 0.4 1
0.24 0.0002 0.358 0.0209 0.709 0.7086 <.0001 <.0001 <.0001
Parental binge drinking 0.03 0.02 0.03 0.05 0.07 0.07 -0.04 0.04 0.04 0.05 1
0.004 0.011 0.003 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001
Academic score -0.07 -0.08 -0.04 -0.12 -0.16 -0.15 0.08 -0.03 0.19 0.11 -0.03 1
<.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 0.002 <.0001 <.0001 0.0004
School Stress 0.29 0.49 0.18 0.05 0.03 0.01 -0.05 0.02 0.004 0.003 0.005 -0.08 1
<.0001 <.0001 <.0001 <.0001 0.0004 0.1476 <.0001 0.043 0.6206 0.727 0.581 <.0001
School Stress

47

48

Table 2. Summary of interested variables in the analysis among youth by gender

Females Males

Variables
Mean SD Mean SD
p
N 7325 6916
Age (year) 14.98 1.72 14.99 1.72 ns
Hostility 2.42 0.50 2.28 2.28 <.0001
Perceived stress 2.60 0.86 2.44 0.89 <.0001
Depression 1.58 0.70 1.46 0.68 <.0001
Parental Monitoring 2.61 0.42 2.48 0.43 <.0001
Latchkey status 1.70 0.85 1.66 .87 <.0.01
Parenting Style (%) <.0.01
Authoritarian 55.49 58.13
Permissive 44.51 41.87
Academic performance 3.54 1.06 3.29 1.17 <.0001
School Stress 2.64 2.49 <.0.01
Parental Heavy drinking 1.96 1.51 2.03 1.57 <.0.01
Light alcohol use (%) 38.03 41.44 0.0004
Moderate alcohol use (%) 23.93 31.47 <.0001
Heavy alcohol use (%) 8.57 23.58 <.0001
Puberty status (%) 91.39 85.76 <.0001
Grade (%)
7
th
19.60 21.85
8
th
20.85 21.99
10
th
29.73 28.76
11
th
29.82 27.39
0.0004
Municipal District (%)
Poor 33.23 31.72
Medium 34.58 35.12
Good 32.19 33.16
ns
School Rank (%)
Poor 25.20 25.49
Medium 27.70 28.43
Good 28.91 30.13
Professional/Vocational 18.18 15.95
0.005
City (%)
ChengDu 15.58 15.14
HangZhou 13.23 13.37
Harbin 12.96 12.87
Kunming 12.96 14.99
QingDao 14.72 13.95
ShenYang 16.14 16.05
Wuhan 14.43 13.62
0.03

49

Results
Characteristics of the sample
A total of 14,434 (93.0%) middle and high school students completed the survey,
and 25,300 parents/guardians (12,445 males and 12,855 females) of the students completed
the adult survey. This analysis included non-missing data from all students in the Youth
survey and the Adult survey. Among the 14,434 students who took part in the student
survey, 11052 students were retained for this analysis.
Correlation analyses [Table 1] indicated very small but significant correlations
between each of the intrapersonal variables (depression (r= 0.11, p<.0001), stress (r= 0.12,
p<.0001 ), hostility (r= 0.06, p<.0001)) and lifetime alcohol use (see methods section below
for variable coding). The correlations were in the expected directions with higher levels of
depression, stress and hostility being correlated with a higher indication of use. Significant
correlations were also found parental monitoring (r=- 0.16, p<.0001), latchkey status (r=
0.05, p<.0001) and lifetime alcohol use. As expected latchkey status was positively
correlated with lifetime alcohol use, and parental monitoring was negatively correlated to
lifetime alcohol use. Correlations in the expected direction were also found for both school
domain variables; school stress (r= 0.05, p<.0001) was positively correlated with lifetime
alcohol use experimentation while academic achievement (r= -0.12, p<.0001) was negatively
correlated with lifetime alcohol use.
Significant differences between genders were found for each variable in all three
domains [Table 2]. Females reported perceiving more stress, depression and hostility than
their male counterparts. Females also reported higher levels of school stress and reported
fairing better academically than males. In the home domain females reported being

50

monitored more closely by their parents and also reported being left in self-care for greater
amounts than males. In addition, more females reported having permissive parents than
males. Although the likelihood of ever drinking was similar for males and females, binge
drinking rates were quite different. When asked if they had consumed 5 or more drinks of
alcohol within any single two hour period during the past 30 days, 24% of males and 8.57 of
females said yes.

Table 3. Odds Ratio of Increasing Levels of AU by Predictors, Females

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
‡
: Life use: 1 or more drinks of alcohol during lifetime;
§
: Monthly use: 1 or more drinks of alcohol during the past 30 days.
||
: Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30 days.
¶
Smoking Behavior: Lifetime smoking used for life AU associations; 30 day smoking used with Monthly AU and Binge drinking associations.

Females N = 7388
Non Alcohol Users N = 3639 Odds Ratio Estimates
*

Life use
‡
Monthly use
§
Binge drinking
||

Effect
OR
95% Confidence
Limits OR
95% Confidence
Limits OR
95% Confidence
Limits
Depression (log) 1.17 (1.09 1.26) 1.27 (1.17 1.37) 1.56 (1.39 1.74)
Psychosocial Stress ns ns ns ns ns ns
Hostility ns ns ns ns ns ns
Exercise (log) ns ns 1.08 (1.01 1.16) 1.32 (1.19 1.48)
Junk food use 1.15 (1.08 1.23) 1.23 (1.14 1.33) ns ns
Intrapersonal Lifestyle Habits Appearance 1.10 (1.03 1.17) 1.12 (1.04 1.21) 1.12 (1.00 1.26)
Health Status (log) ns ns ns ns 1.22 (1.1 1.35)
Smoking behavior
¶
1.56 (1.45 1.67) 1.45 (1.35 1.56) 1.57 (1.45 1.70)
Leisure Time TV time (log) ns ns ns ns ns ns
Parent Monitoring 0.87 (0.81 0.93) 0.90 (0.84 0.98) 0.88 (0.78 0.98)
Parenting Latchkey (log) 1.07 (1.00 1.13) 1.07 (1.00 1.15) ns ns
Parenting Style 0.98 (0.92 1.05) 0.94 (0.87 1.01) ns ns
Academic Score 0.91 (0.85 0.98) 0.85 (0.79 0.93) ns ns
School School Confidence
(log)
ns ns 0.91 (0.85 0.97) 0.88 (0.80 0.96)
School Stress ns ns ns ns 0.89 (0.79 1.00)
Environmental Study time use ns ns ns ns 0.84 (0.75 0.95)
Economic Allowance 1.08 (1.01 1.16) 1.08 (1.00 1.16) 1.25 (1.15 1.37)
Parent drinking (log) 1.13 (1.06 1.20) 1.17 (1.09 1.26) 1.24 (1.10 1.39)
Parental Parent income ns ns ns ns 1.09 (1 1.20)
Parent education ns ns ns ns 1.33 (1.18 1.50)
51

Table 4. Odds Ratio of Increasing Levels of AU by Predictors, Males

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
‡
: Life use: 1 or more drinks of alcohol during lifetime;
§
: Monthly use: 1 or more drinks of alcohol during the past 30 days.
||
: Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30 days.
¶
Smoking Behavior: Lifetime smoking used for life AU associations; 30 day smoking used with Monthly AU and Binge drinking associations.
Males N = 6993
Non Alcohol Users N = 2811 Odds Ratio Estimates
*

Life use
‡
Monthly use
§
Binge drinking
||

Effect
OR
95% Confidence
Limits OR
95% Confidence
Limits OR
95%
Confidence
Limits
Depression (log) ns ns 1.14 (1.06 1.24) 1.35 (1.23 1.47)
Psychosocial Stress ns ns ns ns ns ns
Hostility ns ns ns ns ns ns
Exercise (log) 1.11 (1.04 1.18) ns ns ns ns
Junk food use 1.17 (1.09 1.26) 1.30 (1.20 1.40) 1.36 (1.24 1.49)
Intrapersonal Lifestyle Habits Appearance ns ns 1.11 (1.03 1.19) 1.12 (1.03 1.22)
Health Status (log) ns ns ns ns 1.22 (1.12 1.33)
Smoking behavior
¶
1.55 (1.45 1.65) 1.40 (1.33 1.48) 1.56 (1.48 1.65)
Leisure Time TV time (log) 1.10 (1.03 1.18) 1.10 (1.02 1.18) 1.12 (1.02 1.22)
Parent Monitoring 0.92 (0.85 0.99) 0.92 (0.85 1 ) ns ns
Parenting Latchkey (log) ns ns 1.09 (1.02 1.16) 1.13 (1.04 1.22)
Parenting Style ns ns 0.93 (0.87 1 ) 0.85 (0.79 0.92)
Academic Score ns ns 0.93 (0.87 1.00) 0.86 (0.79 0.94)
School School Confidence (log) 0.92 (0.85 1.00) ns ns ns ns
School Stress ns ns ns ns ns ns
Environmental Study time use ns ns 0.88 (0.82 0.95) 0.82 (0.75 0.91)
Economic Allowance 1.13 (1.04 1.22) 1.11 (1.03 1.19) 1.16 (1.08 1.25)
Parent drinking (log) ns ns 1.12 (1.04 1.20) 1.13 (1.04 1.23)
Parental Parent income ns ns ns ns ns ns
Parent education ns ns ns ns 0.89 (0.80 0.98)
52

53

Model 1: General Alcohol Use Model
This model evaluated the relative odds of our variables of interest predicting a level
of alcohol use between those who endorsed a level of use and those who did not. Tables 3
and 4 report the results for the general alcohol use model analyses.
Intrapersonal
Psychosocial
As expected, results indicated a gradual increase in statistical association of
depression with increasing levels of alcohol use in both genders. In females higher levels of
depression were found to be significantly associated with light use (OR=1.17, 95% CI= 1.09
- 1.26), moderate use (OR= 1.27, 95% CI= 1.17 - 1.37), and binge drinking (OR=1.56, 95%
CI= 1.39 - 1.74). In males higher levels of depression were found to be significantly
associated with moderate use (OR=1.14, 95% CI=1.06 - 1.24), and binge drinking
(OR=1.35, 95% CI=1.23 - 1.47). Contrary to my hypotheses and literature in the field, stress
and hostility were not found to be significantly associated with any level of alcohol use in
neither gender. This may be due to the analyses of environmental variables in the same
model. Indeed one objective of this study was to analyze the effects of both domain
variables on alcohol use while controlling for one another in the same model.

Lifestyle Habits
As expected females with higher levels of exercise were associated with binge
drinking (OR=1.32, 95% CI=1.19 - 1.48). These findings reflect those reported in Western
populations. Also on the part of females, higher levels of junk food consumption were
associated with light use (OR=1.15, 95% CI=1.08 - 1.23), and moderate use (OR=1.23, 95%

54

CI=1.14 - 1.33) but not binge drinking. The association of a specific poor dietary choice
with alcohol use in a Chinese population is a novel finding in the literature.
In males there was a gradual increase in statistical association of higher junk food
use with each level of alcohol use. Junk food was significantly associated with light use
(OR=1.17, 95% CI=1.09 - 1.26), moderate use (OR=1.30, 95% CI=1.20 - 1.40) and binge
drinking (OR=1.36, 95% CI=1.24 - 1.49).
Interestingly, better indices of general health were found to be significantly
associated with binge drinking in females (OR=1.22, 95% CI=1.1 - 1.35) and males
(OR=1.22, 95% CI=1.12 - 1.33). This finding seems to be peculiar to this population. Poor
indices of health have been associated with higher levels of alcohol use in the West. As
expected per co-morbidity models of alcohol and tobacco use in the literature, lifetime
smoking was strongly associated with lifetime alcohol use in both females (OR=1.56, 95%
CI=1.45 - 1.67) and males (OR=1.55, 95% CI=1.45 - 1.65). In females thirty day smoking
was associated with monthly use (OR=1.45, 95% CI=1.34 - 1.56) and binge drinking
(OR=1.57, 95% CI=1.45 - 1.70). Likewise, thirty day smoking was associated with monthly
use (OR=1.40, 95% CI=1.33 - 1.48) and binge drinking (OR=1.56, 95% CI=1.48 - 1.65) in
males.
Environmental
Parenting
Higher levels of parental monitoring among females was negatively associated with
light use (OR=0.87, 95% CI=0.81 - 0.93). In males, higher levels of latchkey status were
associated binge drinking (OR=1.13, 95% CI=1.04 - 1.22). Also among males, authoritarian
parenting styles were associated with lower levels of binge drinking (OR=0.86, 95%

55

CI=0.79 - 0.94). Investigations looking at the effects of parenting on youth alcohol use have
consistently found parental monitoring and latchkey status to be highly associated with
youth alcohol use behaviors in both genders. As with the findings reported for the
psychosocial variables, it may be the case that the effects of parenting are diminished when
taking into account intrapersonal variables in the same model.
School
As expected higher academic scores were associated with lower moderate use
(OR=0.85, 95% CI=0.79 - 0.93) in females, and with lower binge drinking (OR=0.86, 95%
CI=0.79 - 0.94) in males. More time spent on school work after school was related to lower
binge drinking levels (OR=0.84, 95% CI=0.75 - 0.95) in females, and lower levels of binge
drinking (OR=0.82, 95% CI=0.75 - 0.91) in males. These results reflect those found in
Western populations in which higher academic scores are associated with lower alcohol use.
The protective effect of more time spent on school work after school for alcohol use is a
novel finding in alcohol literature among this population.
Economic
More disposable income in the form of allowance was found to be associated with
greater alcohol use in both genders. Greater disposable income was associated with binge
drinking (OR=1.25, 95% CI=1.15 - 1.37) in females; and with increasing levels of binge
drinking (OR=1.16, 95% CI=1.08 - 1.25) in males. Indeed, this is a finding that has been
replicated across many different populations. A number of studies have reported that the
cost of alcoholic beverages is many times prohibitive to its use in youth populations.

56

Parental
Parental drinking was associated with incremental levels of alcohol use in females.
Higher parental drinking was associated with higher levels of light use (OR=1.13, 95%
CI=1.06 - 1.20), moderate use (OR=1.17, 95% CI=1.09 - 1.26), and binge drinking
(OR=1.24, 95% CI=1.10 - 1.39) in females. Higher degree of parental education was
positively associated with binge drinking (OR=1.33, 95% CI=1.18 - 1.50) in females and
negatively associated with binge drinking (OR=0.89, 95% CI=0.80 - 0.98) in males. This
finding reflects what is postulated by the multistage social learning model of substance use;
adolescents may become involved with greater uses of alcohol if they observe their parents
engaging in alcohol use (Simons, Conger et al. 1988; Petraitis, Flay et al. 1995).

Table 5. Odds Ratio of Exclusive Levels of AU by Predictors of Interest, Females

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
‡
: Life use: 1 or more drinks of alcohol during lifetime but not in the last 30 days;
§
: Monthly use: 1 or more drinks of alcohol during the past 30 days but not heavy use.
||
: Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30 days.
¶
Smoking Behavior: Lifetime smoking used for life AU associations; 30 day smoking used with Monthly AU and Binge drinking associations.

Females
Non Alcohol Users N = 3639 Odds Ratio Estimates
*

Life use
‡
vs. Non-use
Monthly use
§
vs. Non-use
Binge drinking
||
vs. Non-use
N = 2232 N = 1140 N = 338

Effect
OR
95% Confidence
Limits OR
95% Confidence
Limits OR
95% Confidence
Limits
Depression (log) 1.12 (1.03 1.22) 1.24 (1.13 1.37) 1.57 (1.37 1.8 )
Psychosocial Stress ns ns ns ns 0.75 (0.64 0.88)
Hostility ns ns ns ns 1.28 (1.12 1.46)
Exercise (log) ns ns ns ns 1.22 (1.08 1.38)
Junk food use 1.11 (1.03 1.19) 1.31 (1.20 1.44) 1.37 (1.21 1.55)
Intrapersonal
Lifestyle
Habits Appearance
1.07 (1.00 1.16) 1.14 (1.04 1.24) 1.50 (1.31 1.72)
Health Status (log) ns ns ns ns 1.27 (1.12 1.43)
Smoking behavior
¶
1.43 (1.33 1.55) 1.43 (1.29 1.57) 2.05 (1.84 2.29)
Leisure Time TV time (log) ns ns ns ns ns ns
Parent Monitoring 0.87 (0.80 0.94) 0.82 (0.75 0.91) 0.84 (0.73 0.95)
Parenting Latchkey (log) ns ns 1.09 (1.01 1.19) 1.23 (1.09 1.39)
Parenting Style ns ns ns ns ns ns
Academic Score 0.92 (0.85 1.00) 0.81 (0.74 0.89) ns ns
School School Confidence
(log)
ns ns
0.92 (0.85 1.00) 0.81 (0.73 0.90)
Environmental Economic Allowance ns ns ns ns 1.33 (1.20 1.48)
Parent drinking (log) 1.08 (1.01 1.15) 1.18 (1.09 1.29) 1.39 (1.22 1.6 )
Parental Parent income ns ns ns ns 1.11 (1.00 1.24)
Parent education ns ns ns ns ns ns
57

Table 6. Odds Ratio of Exclusive Levels of AU by Predictors of Interest, Males

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
‡
: Life use: 1 or more drinks of alcohol during lifetime but not in the last 30 days.
§
: Monthly use: 1 or more drinks of alcohol during the past 30 days but not heavy use.
||
: Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30 days.
¶
Smoking Behavior: Lifetime smoking used for life AU associations; 30 day smoking used with Monthly AU and Binge drinking associations.

Males
Non Alcohol Users N = 2811 Odds Ratio Estimates
*

Life use
‡
vs. Non-use
Monthly use
§
vs. Non-use
Binge drinking
||
vs. Non-use
N = 1986 N = 1292 N = 860

Effect
OR
95% Confidence
Limits OR
95% Confidence
Limits OR
95% Confidence
Limits
Depression (log) 0.88 (0.80 0.97) ns ns 1.15 (1.03 1.29)
Psychosocial Stress ns ns ns ns 1.17 (1.02 1.34)
Hostility ns ns ns ns ns ns
Exercise (log) 1.10 (1.02 1.19) 1.08 (0.99 1.18) ns ns
Junk food use ns ns 1.27 (1.16 1.40) 1.42 (1.27 1.59)
Intrapersonal Lifestyle Habits Appearance ns ns ns ns ns ns
Health Status (log) ns ns ns ns 1.22 (1.09 1.36)
Smoking behavior
¶
1.42 (1.31 1.53) 1.38 (1.27 1.49) 1.89 (1.75 2.05)
Leisure Time TV time (log) 1.07 (0.99 1.17) 1.12 (1.02 1.23) 1.27 (1.13 1.43)
Parent Monitoring 0.91 (0.83 0.99) 0.86 (0.78 0.95) 0.86 (0.76 0.97)
Parenting Parenting Style ns ns ns ns 0.85 (0.77 0.94)
Academic Score ns ns ns ns 0.86 (0.77 0.96)
School School Confidence (log) ns ns ns ns 0.88 (0.78 0.99)
School Stress ns ns 1.15 (1.04 1.28) ns ns
Environmental Study time use ns ns 0.91 (0.83 1.00) 0.79 (0.70 0.90)
Economic Allowance 1.10 (1.01 1.20) 1.09 (0.99 1.20) 1.19 (1.06 1.32)
Parent drinking (log) ns ns ns ns 1.13 (1.01 1.26)
Parental Parent income ns ns ns ns 1.25 (1.11 1.39)
Parent education ns ns ns ns ns ns
58

59

Model 2 Exclusive alcohol use model
This model evaluated the relative odds of our variables of interest predicting a level
of alcohol use between those who endorsed a single level of use and non alcohol users.
Tables 5 and 6 report the results for the exclusive alcohol use model analyses.
Intrapersonal
Psychosocial
Results again indicated a gradual increase in statistical association of depression
with increasing levels of alcohol use in females. Females with higher levels of depression
were found to be at higher odds for light use (OR=1.12, 95% CI= 1.03 - 1.22), moderate use
(OR= 1.24, 95% CI= 1.13 - 1.37), and binge drinking (OR=1.57, 95% CI= 1.37 - 1.80).
Males with higher levels of depression were found to be at lower odds for light use
(OR=0.88, 95% CI=0.80 - 0.97), but at higher odds for binge drinking (OR=1.15, 95%
CI=1.03 - 1.29). High levels of stress lowered the odds with binge drinking (OR=0.75, 95%
CI=0.64 - 0.88) in females, and increased the odds associated with binge drinking in males
(OR=1.17, 95% CI=1.02 - 1.34). In females higher indices of hostility were associated with
higher odds for binge drinking (OR=1.28, 95% CI=1.12 - 1.46). In this model of analysis
we observe that stress and hostility are significantly associated with selected alcohol uses in
both genders. This marks a difference with results reported in the first analysis model where
no associations were found between stress, hostility and alcohol use in neither gender.

Lifestyle Habits
As in model 1 females with higher levels of exercise were associated with higher
odds for binge drinking (OR=1.22, 95% CI=1.08 - 1.38). Females with greater junk food

60

consumption were incrementally at higher odds for moderate use (OR=1.31, 95% CI=1.20 -
1.44), and binge drinking (OR=1.37, 95% CI=1.21 - 1.55).
Likewise males with higher levels of junk food consumption were incrementally at
greater odds for moderate use (OR=1.27, 95% CI=1.16 - 1.40) and binge drinking
(OR=1.42, 95% CI=1.27 - 1.59). For females the greater importance given toward
appearance the greater the odds for binge drinking (OR=1.50, 95% CI=1.31 - 1.72). Again,
surprisingly, better indices of general health were found to be associated with higher odds
for binge drinking (OR=1.27, 95% CI=1.12 - 1.43) in females and males (OR=1.22, 95%
CI=1.09 - 1.36).
In females lifetime smoking was associated with a greater odds of lifetime use
(OR=1.43, 95% CI=1.33 - 1.55). An indication of 30 day smoking was associated with a
greater odds of monthly use (OR=1.43, 95% CI=1.29 - 1.57), and incrementally with binge
drinking (OR=2.05, 95% CI=1.84 - 2.29). Positive smoking behavior in males was
associated with greater odds for lifetime use (OR=1.42, 95% CI=1.31 - 1.53), monthly use
(OR=1.38, 95% CI=1.27 - 1.49), and binge drinking (OR=1.89, 95% CI=1.75 - 2.05). These
findings for lifestyle habits show that many of the significant associations found in the first
analysis model continue to be significant when restricting the analysis to those who endorsed
a single level of use and non alcohol users.
Leisure Time
As in the first model of analysis, more amounts of time spent watching television
among males was associated with a higher odds of binge drinking (OR=1.27, 95% CI=1.13 -
1.43). This finding supports those reported in Western populations. It is important to point
out that it is not currently known what types of TV shows are watched by these males.

61

Investigations conducted in the West have found that Western television programs contain a
high number of positive drinking scenes and that high exposure to such programming is
associated with youth alcohol use.
Environmental
Parenting
Higher levels of parental monitoring were associated with a lower odds of light use
(OR=0.87, 95% CI=0.80 - 0.94), moderate use (OR=0.82, 95% CI=0.75 - 0.91), and binge
drinking (OR=0.84, 95% CI=0.73 - 0.95) in females. In males higher levels of parental
monitoring were associated with a lower odds of and moderate use (OR=0.86, 95% CI=0.78
- 0.95), and binge drinking (OR=0.86, 95% CI=0.76 - 0.97). Females with higher levels of
latchkey status were at greater odds for binge drinking (OR=1.23, 95% CI=1.09 - 1.39).
Authoritarian parenting styles were associated with lower odds for binge drinking (OR=0.86,
95% CI=0.76 - 0.97) among males. As in the first model of analysis these findings show
that not all parenting variables were found to be significantly associated with youth alcohol
use as is reported in the literature. Again, this may be due the effect of analyzing both
intrapersonal and environmental domain factors in the same model of analysis.
School
Females with higher academic scores were at lower odds for moderate use
(OR=0.81, 95% CI=0.74 - 0.89). Males with higher academic scores were at lower odds for
binge drinking (OR=0.86, 95% CI=0.76 - 0.97). Higher levels of school confidence were
associated with lower odds for binge drinking (OR=0.81, 95% CI=0.73- 0.90) in females;
and males (OR=0.88, 95% CI=0.78- 0.99). More time spent on school work after school put
males at lower odds for binge drinking (OR=0.79, 95% CI=0.70 - 0.90). These findings are

62

consistent with those reported in the first model of analysis. We see that these findings hold
when restricting the analysis to those who endorsed a single level of use and non alcohol
users.
Economic
Similarly to findings reported in the first model of analysis, females with greater
allowance were at greater odds for binge drinking (OR=1.33, 95% CI=1.20 - 1.48).
Likewise males with greater allowance were at greater odds for binge drinking (OR=1.19,
95% CI=1.06 - 1.32).
Parental
In females higher parental drinking was associated incrementally with a higher odds
for moderate use (OR=1.18, 95% CI=1.09 - 1.29), and binge drinking (OR=1.39, 95%
CI=1.22 - 1.60). Higher parental income was associated with a higher odds for binge
drinking in males (OR=1.25, 95% CI=1.11- 1.39). These findings reflect those from the first
level of analysis and those reported in Western populations.
Model 3: Incremental alcohol use model
Two models were tested, one in which the relative odds of monthly alcohol use
among lifetime alcohol users was assessed and another in which the relative odds of binge
drinking among monthly alcohol users was assessed. Tables 7 and 8 report the results for the
relative odds of monthly alcohol use among lifetime alcohol users. Tables 10 and 11 report
the results for the relative odds binge drinking among monthly alcohol users.

63

Table 7. Relative Odds of Monthly Alcohol Use Among Female Lifetime Users

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
†
: Standardized scores used for predictors of interest.

Female lifetime users were at a greater odds for monthly alcohol use if they had
higher consumption of junk food (OR=1.17, 95% CI=1.06 - 1.29), if they had smoked in the
past 30 days (OR=1.28, 95% CI=1.15 - 1.42), and if their parents were monthly users
(OR=1.16, 95% CI=1.05 - 1.27). Female lifetime users were at a lower odds for monthly
use if they had higher academic scores (OR=0.87, 95% CI=0.78 - 0.97).

Females N=
3710 Odds Ratio Estimates
*

Effect
†
OR
95% Confidence
Limits
Depression (log) ns ns
Psychosocial Stress ns ns
Hostility ns ns
Exercise (log) ns ns
Junk food use 1.17 (1.06 1.29)
Intrapersonal Lifestyle Habits Appearance ns ns
Health Status (log) ns ns
Smoking behavior
¶
1.28 (1.15 1.42)
Leisure Time TV time (log) ns ns
Parent Monitoring ns ns
Parenting Latchkey (log) ns ns
Parenting Style ns ns
Academic Score 0.87 (0.78 0.97)
School School Confidence (log) 0.92 (0.83 1.00)
School Stress ns ns
Environmental Study time use ns ns
Economic Allowance ns ns
Parent drinking (log) 1.16 (1.05 1.27)
Parental Parent income ns ns
Parent education ns ns

64

Table 8. Relative Odds of Monthly Alcohol Use Among Male Lifetime Alcohol Users

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
†
: Standardized scores used for predictors of interest.

Male lifetime users were at a greater odds for monthly alcohol use if they had higher
consumption of junk food (OR=1.20, 95% CI=1.09 - 1.33), were more concerned about their
appearance (OR=1.14, 95% CI=1.03 - 1.25), if they had smoked in the past 30 days
(OR=1.15, 95% CI=1.06 - 1.24), and if their parents were monthly users (OR=1.11, 95%
CI=1.01 - 1.22).

Males N = 4138 Odds Ratio Estimates
*

Effect
†
OR
95% Confidence
Limits
Depression (log) ns ns
Psychosocial Stress ns ns
Hostility ns ns
Exercise (log) ns ns
Junk food use 1.20 (1.09 1.33)
Intrapersonal Lifestyle Habits Appearance 1.14 (1.03 1.25)
Health Status (log) 1.01 (0.91 1.11)
Smoking behavior
¶
1.15 (1.06 1.24)
Leisure Time TV time (log) ns ns
Parent Monitoring ns ns
Parenting Latchkey (log) ns ns
Parenting Style ns ns
Academic Score ns ns
School School Confidence (log) ns ns
School Stress ns ns
Environmental Study time use ns ns
Economic Allowance ns ns
Parent drinking (log) 1.11 (1.01 1.22)
Parental Parent income ns ns
Parent education ns ns

65

Table 9. Relative Odds of Binge Drinking Among Female Monthly Alcohol Users

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
†
: Standardized scores used for predictors of interest.

Female monthly alcohol users were at greater odds for binge drinking if they had
higher levels of depression (OR=1.36, 95% CI=1.12 - 1.64), if they had smoked in the past
30 days (OR=1.19, 95% CI=1.05 - 1.37), if they had higher allowances (OR=1.41, 95%
CI=1.19 - 1.67), if their parents were monthly users (OR=1.20, 95% CI=1.00 - 1.43).

Females N =
1478 Odds Ratio Estimates
*

Effect
†
OR
95% Confidence
Limits
Depression (log) 1.36 (1.12 1.64)
Psychosocial Stress ns ns
Hostility ns ns
Exercise (log) ns ns
Junk food use ns ns
Intrapersonal Lifestyle Habits Appearance ns ns
Health Status (log) ns ns
30 day smoking 1.19 (1.05 1.37)
Leisure Time TV time (log) ns ns
Parent Monitoring ns ns
Parenting Latchkey (log) ns ns
Parenting Style ns ns
Academic Score ns ns
School School Confidence (log) ns ns
School Stress ns ns
Environmental Study time use ns ns
Economic Allowance 1.41 (1.19 1.67)
Parent drinking (log) 1.2 (1.00 1.43)
Parental Parent income ns ns
Parent education ns ns

66

Table 10. Relative Odds of Binge Drinking Among Male Monthly Alcohol Users

*
: Adjusted for city, academic level of school attended, economic level of municipal district.
†
: Standardized scores used for predictors of interest.

Male monthly alcohol users were at greater odds for binge drinking if they had
higher levels of depression (OR=1.42, 95% CI=1.23 - 1.63), if they had higher indices of
health status (OR=1.16, 95% CI=1.02 - 1.32), and if they had smoked in the past 30 days
(OR=1.32, 95% CI=1.21 - 1.43). Male monthly users were at lower odds for binge drinking
if they had higher academic scores (OR=0.80, 95% CI=0.70 - 0.91), and spent more time on
school work after school (OR=0.85, 95% CI=0.74 - 0.97).

Males N =
2152 Odds Ratio Estimates
*

Effect
†
OR
95% Confidence
Limits
Depression (log) 1.42 (1.23 1.63)
Psychosocial Stress ns ns
Hostility ns ns
Exercise (log) ns ns
Junk food use ns ns
Intrapersonal Lifestyle Habits Appearance ns ns
Health Status (log) 1.16 (1.02 1.32)
30 day smoking 1.32 (1.21 1.43)
Leisure Time TV time (log) ns ns
Parent Monitoring ns ns
Parenting Latchkey (log) ns ns
Parenting Style ns ns
Academic Score 0.80 (0.70 0.91)
School School Confidence (log) ns ns
School Stress ns ns
Environmental Study time use 0.85 (0.74 0.97)
Economic Allowance ns ns
Parent drinking (log) ns ns
Parental Parent income ns ns
Parent education ns ns

67

Discussion
Female Results Pattern
As expected, depression was found to be incrementally associated with all three
levels of alcohol use in both the general and exclusive analysis models. In addition
depression was associated with odds of binge drinking among monthly alcohol users. These
results replicate findings that associate depression with alcohol use in Western populations.
However, the results presented here have contributed the novel findings that depression was
increasingly associated with incremental levels of alcohol use. The fact that this has been
demonstrated in a general alcohol analysis model and in an alcohol exclusive model; and
that high levels of depression have been shown to increase the odds for binge drinking
among female monthly alcohol users serves to further solidify this contribution to the field
of youth alcohol use research
Higher levels of stress were shown to be associated with lower odds for binge
drinking, and higher levels of hostility were associated with greater odds for binge drinking
in the alcohol exclusive model. Nonetheless these results did not support my hypotheses of
finding increasing associations of stress and hostility with incremental levels of alcohol use.
While findings from other studies have reported associations of hostility and to a lesser
degree stress with youth alcohol use behaviors, these studies tend to focus exclusively on
intrapersonal domain factors. It may be the case that my findings for stress and hostility
were not supported given the fact that a number of environmental domain variables were
present in the model. Indeed, one of the purposes of this study was to analyze the effects of
intrapersonal and environmental domain variables on alcohol use behaviors while tested in
the same model. Further research is needed which can clarify the interactions that

68

intrapersonal variables have on environmental variables (and vice versa) in their association
with youth alcohol use behaviors.
A high concern over appearance was found to be associated with all three levels of
alcohol use in both the general and exclusive analysis models, and was associated with
incremental levels of alcohol use in the exclusive analysis only. These findings support
findings in the literature that indicate that concern for body image is an important factor the
onset and maintenance of substance abuse among adolescents (Guinn B 1997; Nieri, Kulis et
al. 2005). In this case substance use may be acting as a coping strategy for adolescents who
have indicated body image problems (Guinn B 1997; Nieri, Kulis et al. 2005). Findings
from other studies have linked poor body image to low self-esteem and depression, which
are risk factors for substance use (Guinn B 1997; Nieri, Kulis et al. 2005). A novel
contribution to the literature of adolescent alcohol use are the associations that have been
presented here between concern over appearance and alcohol use behaviors in an adolescent
female Chinese population. Further research is required which can further explore other risk
factors associated with concern over appearance among Chinese adolescent females.
As expected smoking behavior was associated with all three levels of alcohol use in
both the general and exclusive analysis models and was associated with incremental levels of
alcohol use in the exclusive analysis only. Smoking behavior was also found to be
associated with odds of monthly alcohol use among lifetime users and with odds of binge
drinking among monthly alcohol users. These results confirm the findings that indicate the
existence of a strong co-morbidity between smoking behaviors and drinking in late
adolescence and adults. It is alarming that these findings are present in adolescents as young
as this study population.

69

Junk food use was associated with all three levels of alcohol use in the exclusive
analysis model only. Additionally junk food use was associated with odds of monthly
alcohol use among lifetime users. The results for junk food use have proven very interesting
to this investigator. Even though findings in the literature indicate an association between
poor dietary choices and alcohol use, the fact that this study has associated a specific youth
poor dietary practice to alcohol use behaviors presents a novel contribution to the field.
Further research is needed which can establish if behaviors such as poor dietary choices,
such as junk food use, form part of an “indulgence” model of use behaviors.
While parental monitoring was found to be associated with all three levels of alcohol
use in both the general and exclusive analysis models, no consistent findings were present
for latchkey status or parenting styles in females. Again this may be due to the fact that a
number of intrapersonal domain factors were present in the model and these effects were
lost. This warrants further research given the fact that a previous study by this investigator
found latchkey status and parenting style to be highly associated with alcohol use behaviors
in this population.
High academic scores proved to be the variable that was more strongly associated
with lower odds for use behaviors. Higher academic scores were associated with lower odds
for monthly use in both analysis models. These results support findings in the West that
report high academic achievement to be associated with lower alcohol use in adolescents.
The fact that Chinese culture emphasizes high academic achievement in its youth must be
seen as an important opportunity for prevention scientists to further explore the role of
academic achievement in intervention efforts that aim to delay onset and stop transition to
heavier forms of use in this population.

70

As expected and following elements of the multistage social learning model
(Simons, Conger et al. 1988) Parental drinking was found to be incrementally associated
with all three levels of alcohol use in both the general and exclusive analysis models.
Parental drinking was also associated with odds of monthly alcohol use among lifetime users
and with odds of binge drinking among monthly alcohol users.
Male Results Pattern
Higher levels of depression were associated with greater odds for binge drinking in
the general alcohol analysis model. Nonetheless no hypotheses were supported for any
psychosocial domain variable association with incremental alcohol use. As discussed
previously, a possible contribution of this study is presenting the loss of association of stress
and hostility with alcohol use behaviors when environmental variables are present.
As with females, smoking behavior was found to be incrementally associated with
all three levels of alcohol use in both the general and exclusive analysis models. Smoking
behavior was also associated with odds of monthly alcohol use among lifetime users and
with odds of binge drinking among monthly alcohol users. These findings indicate that the
co-morbidity model of smoking and alcohol use behaviors is supported in both genders
among Chinese adolescents. This shows that empirical models attempting to explain risk
factors for alcohol use among Chinese adolescents must include smoking behaviors as
important determinants for both experimental use and for heavier uses such as binge
drinking behaviors.
As with females junk food use was observed to be associated with incremental levels
of alcohol use in both analysis models. Junk food use was also associated with odds of
monthly alcohol use among lifetime users. Coupled with the results presented among

71

females these results present a novel contribution to youth alcohol use research. This study
is among the first to associate a specific poor adolescent dietary choice to alcohol use
behaviors. These findings indicate that it is important to take this variable into account when
analyzing general risk factors for alcohol use in this population. Further research is needed
which can clarify if this behavior forms part of a more general model of “indulgent
behaviors”.
Among males we found that amount of time spent watching TV was associated with
all three levels of alcohol use in the general analysis model only. Further investigations are
necessary which can identify the reasons for such association. As has been discussed
content analyses of alcohol use scenes on television suggest that incidences of drinking
occur frequently, and that these portrayals generally present drinking as an activity that is
associated with happiness, social achievement, relaxation, and camaraderie (Atkin 1990;
Van Den Bulck and Beullens 2005). It will be important for future studies to conduct
content analysis of Chinese television programming in order to identify what is driving the
association between amount of time spent watching TV and increased alcohol use.
My hypotheses regarding parenting variables were not supported. Further
investigations are required which analyze possible interactions between intrapersonal and
environmental variables in their association with alcohol use behaviors to see if more refined
analyses are required when variables from both domain factors are present in the same
model. More detailed analyses of this sort can help direct investigators in their efforts of
analyzing intrapersonal and environmental influences of adolescent alcohol use.
Investigators must be able to determine if it is necessary to analyze multiple domains when
analyzing youth AU determinants. This investigation has shown that indeed results may

72

vary between studies that analyze both intrapersonal and environmental domains in a single
model and those that investigate these domains exclusively.
As we have indicated previously there have been abundant youth AU studies
conducted in the West and few conducted in Asian countries where drinking patterns and
social/cultural norms are different (Janghorbani, Ho et al. 2003; Hao, Chen et al. 2005; Xing,
Ji et al. 2006; Zhou, Su et al. 2006). The influences of intrapersonal and environmental
factors have rarely been investigated in this population. The results presented here show that
while many alcohol use determinants are similar to those found in the West, some seem to be
particular to Chinese youth populations. These differences need to be explored further and if
possible linked to investigations of alcohol biological predisposition in order to fully
differentiate socio-cultural determinants from biological ones.
Effective youth alcohol use prevention efforts must consider the potential
contributors to this behavior from the most salient domains of an adolescent’s life. This
study has helped in identifying just such contributors among a population of Chinese
adolescents. The knowledge of the role that potential co-morbid psychosocial disorders may
have on alcohol use behaviors can help direct investigators to focus intervention efforts on
the underlying influences of alcohol use and on concurrent psychosocial conditions that exist
in this population.

73

CHAPTER FIVE: MODERATION EFFECTS BETWEEN INTRAPERSONAL AND
ENVIRONMENTAL DETERMINANTS OF ALCOHOL USE IN CHINESE YOUTH
The first study in this dissertation investigation tested the direct effect of selected
intrapersonal and environmental domain factors on the alcohol use behaviors of a Chinese
youth population. It is important to note that the direct effects of both domain factors were
tested while controlling for the influence of one another in the same statistical analysis
model. This is an important consideration given that the majority of investigations tend to
test the direct effects of either intrapersonal or environmental variables exclusively. Results
from the first study showed that that results may vary between studies that analyze both
intrapersonal and environmental domains in a single model and those that investigate these
domains exclusively.
The information presented in this first study has given us a better understanding of
how concurrent determinants from different life domains affect youth AU behaviors. We
know that AU determinants do not influence consumption behaviors in a vacuum, but rather
that AU behaviors are determined by many different competing factors from different
domains of an adolescent’s life. Indeed, research has shown that substance use results from
an interaction of a number of factors including cognitive, attitudinal, social, personality,
pharmacological, biological, and developmental variables (Mayhew, Flay et al. 2000;
Donovan and . : 2004; Glantz and Mandel 2005; Sher, Grekin et al. 2005).
A review of the pertinent literature has shown that important alcohol use
determinants of youth AU have been identified in both intrapersonal and environmental life
domains (Hawkins and Weis 1985; Hawkins, Catalano et al. 1992). The findings of the
previous study have given us a better understanding of how the selected alcohol use

74

predictors possibly influence adolescent alcohol use behaviors in their daily lives. Results
from the previous study showed that particular intrapersonal and environmental domain
factors lost their significant associations with alcohol use behaviors when variables from
both domain factors were present in the same model. It is worth noting that previous studies
have been conducted by this investigator in which each domain was tested exclusively for
their influence on youth alcohol use. These studies found significant associations of each
intrapersonal and environmental domain factor with similar alcohol use behaviors in this
same population.
To further our understanding of the effects that intrapersonal and environmental
domain predictors have on one another in their association with adolescent AU behaviors we
can investigate the potential interaction effects that these predictors have on one another in
their relationship to youth alcohol use behaviors.
Adolescent alcohol use investigators point to the fact that not much is currently
known about which determinants moderate the relationship between alcohol use risk factors
and alcohol use behaviors (Mayhew, Flay et al. 2000; Donovan and . : 2004; Glantz and
Mandel 2005; Sher, Grekin et al. 2005). The majority of findings of current investigations in
the field provide us with correlates of alcohol use behaviors but rarely with more detailed
information regarding the nature of the associations that exist between identified risk and
protective factors and alcohol use behaviors(Donovan, Jessor et al. 1983; Donovan and
Jessor 1985; Donovan et al. 2004). Analyses are needed which can clarify if associations
between risk factors and AU behaviors differ in magnitude or direction as a function of third
variables (Donovan, Jessor et al. 1983; Donovan and Jessor 1985; Donovan and . : 2004).
This type of information is important in that it can provide more thorough explanations of
the mechanisms that particular alcohol use determinants have on alcohol use onset and

75

progression. It is important to test more complex models of alcohol use determinants which
can help better establish the nature of the inter-relationships that may exist between such
determinants in their association with alcohol use behaviors.
This second study explored the effects that intrapersonal dispositional variables had
on the relationship between environmental factors and alcohol use behaviors. While there is
evidence for the contributing effects of home, school, and parental environments on youth
alcohol use I wanted to investiagte how intrapersonal dispositional variables affect these
relationships. There is a growing need to study the effects of biological and genetic
influences on all forms of health risk behavior. Studies in several countries have
demonstrated that some behavioral characteristics present at childhood are associated with
alcohol-related problems later in adolescence. While biological and genetic analyses are
beyond the scope of this paper, one question this investigation posed is whether dispositional
proxy biological variables such as depression, stress and hostility, changed the magnitude of
association (moderated) between environmental determinants and alcohol use behaviors.
This study explored the moderating effect of each intrapersonal variable on the
relationship between selected environmental variables and each level of alcohol use in youth.
Specifically this study investigated the moderating effects of depression, stress and hostility
on the relationship between the environmental parenting (latchkey status), school (academic
score), economic (allowance), parental demographics (parental drinking, highest education
level, and income) variables and two levels of alcohol use. The two levels of alcohol use
used in this analysis were monthly use, and binge drinking. These environmental domain
variables have been selected because they best represent adolescent environmental

76

influences from the pool of available variables. The analyses were conducted on cross-
sectional data from the CSCS baseline survey.
Hypotheses
(1) Intrapersonal variables (depression, hostility, stress) would moderate the
relationship between latchkey status and each alcohol use consumption level. I
expected that the relationship between each alcohol use level and each intrapersonal
variable would significantly differ by high and low levels of latchkey status.
(2) Intrapersonal variables (depression, hostility, stress) would moderate the
relationship between academic scores and each alcohol use consumption level. I
expected to find that the effect of each intrapersonal variable on each alcohol use
level would be lower for those with high academic scores.
(3) Intrapersonal variables (depression, hostility, stress) would moderate the
relationship between allowance and each alcohol use consumption level. I expected
that the relationship between each alcohol use level and each intrapersonal variable
would significantly differ by high and low levels of allowance received.
(4) Intrapersonal variables (depression, hostility, stress) would moderate the
relationship between parental variables (parental drinking, parental education) and
each alcohol use consumption level. I expected that the relationship between each
alcohol use level and each intrapersonal variable would significantly differ by high
and low levels of each parental variable.
(5) In accordance with multistage social learning model (Simons, Conger et al.
1988) I expected that moderation effects would be stronger in association with binge
drinking behaviors.

77

Analysis
I tested a model of interaction in which I examined the potential moderating effects
of the selected dispositional variables on the relationships between selected environmental
variables and each level of alcohol use. Analyses for this study were performed on cross-
sectional data from China Seven Cities Study (CSCS) baseline survey. Due to the clustering
of students within schools and the possible intra-school correlation between students within
the same school, a generalized linear model was applied in the analysis. In this model, both
inter-school and intra-school correlations were assumed for self-reported measure, and a
logit link function was assumed to transform the dichotomous outcome
This study analyzed the cross-sectional associations of the selected intrapersonal and
environmental variables and alcohol use levels. Regression coefficients of interaction
product variables were evaluated as well. This study analyzed the moderating effects of
depression, stress and hostility on the relationship between the environmental parenting
(latchkey status), school (academic score), economic (allowance), and parental
demographics (parental drinking, highest education level) variables and three levels of
alcohol use (lifetime, monthly, and binge drinking).
As in Study 1 all independent variables were tested in a single model in order to
control for individual effects. Possible confounders adjusted for in the regressions included
age, grade, puberty, city, type and academic level of school attended (poor, medium, good,
professional/vocational), economic level of municipal district (poor, medium, good),
parental education level, parental income level, and parental heavy alcohol use. Analyses
were conducted separately by gender in order to assess gender differences that may exist in
these associations.

78

Three models were tested in order to fully understand the relationships that exist
between our predictors of interest and each level of alcohol use.
In the first model (general alcohol use model) 3 dichotomous levels of alcohol were
tested: lifetime alcohol use (yes/no), monthly alcohol use (yes/no), and binge drinking
(yes/no). Statistical associations between each intrapersonal and environmental variable and
each level of alcohol use were analyzed. This model evaluated the relative odds of the
variables of interest predicting a level of alcohol use between those who endorsed a level of
use and those who did not.
A second model was tested (exclusive alcohol use model) in which exclusive
alcohol use categories were used. Three levels of use were used which were exclusive of one
another. The lifetime use category for example only comprehended those users who
endorsed lifetime use but not monthly use nor binge drinking. This model evaluated the
relative odds of our variables of interest predicting a level of alcohol use between those who
endorsed a single level of use and non alcohol users.
Lastly a third model was tested (incremental alcohol use model) where I analyzed
which variables predicted transition from one level of alcohol use to an incremental level of
use (e.g. life use to monthly use or monthly use to binge drinking). Two models were tested,
one in which the relative odds of monthly alcohol use among lifetime alcohol users was
assessed, and another in which the relative odds of binge drinking among monthly alcohol
users was assessed.

79

Results
Model 1: General Alcohol Use Model
This model evaluated the relative odds of our variables of interest predicting a level of
alcohol use between those who endorsed a level of use and those who did not. Tables 11 and
12 report the results for the general alcohol use model analyses.
Females
Table 11. Environmental and Intrapersonal Variable Interactions, Females

Females N = 7388

Beta Estimates
*

Non Alcohol Users =
3639
Monthly Alcohol
‡
Binge drinking
§

Effect
†
Beta SE P Beta SE P

Parenting
Depression
(log)*Latchkey(log)
0.095 0.037 0.012 ns ns ns

School
Depression
(log)*Academic score
ns ns ns -0.22 0.060 0.0004
Depression
Economic
Depression
(log)*Allowance
ns ns ns 0.171 0.049 0.0005

Depression (log)*Parental
drinking
ns ns ns -0.14 0.058 0.0194

Parental
Depression (log)*Parent
education
ns ns ns 0.261 0.058 <.0001
Parenting Stress*Latchkey(log) ns ns ns ns ns ns
School Stress*Academic score ns ns ns ns ns ns
Stress Economic Stress*Allowance -0.11 0.040 0.007 -0.15 0.049 0.0032
Stress*Parental drinking ns ns ns ns ns ns
Parental Stress*Parent education ns ns ns -0.39 0.064 <.0001
Parenting Hostility*Latchkey(log) ns ns ns ns ns ns
School Hostility*Academic score -0.09 0.040 0.017 0.123 0.054 0.0237
Hostility Economic Hostility*Allowance ns ns ns ns ns ns
Hostility*Parental drinking ns ns ns ns ns ns
Parental Hostility*Parent education ns ns ns ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Monthly use: 1 or more drinks of alcohol during the past 30 days.
§:
Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30
days.

80

Depression
A positive statistically significant interaction was found between depression and
latchkey status in their effect on monthly alcohol use (Std. β = 0.095, p = 0.01). This
indicates that the relationship between monthly alcohol use and depression significantly
differed between those with low latchkey status and those with high latchkey status. The
impact of the two variables together was substantially greater than the additive effect of the
two variables in the model. While this interaction was not in the expected direction results
do indicate that alcohol use varies by amount of time spent in self care. This finding
confirms my general hypothesis that intrapersonal variables moderate the effect between
environmental variables and youth alcohol use.
As expected, a negative statistically significant interaction was found between
depression and academic score in their effect on binge drinking (Std. β = -0.21, p = 0.0004).
The negative interaction meant that the effect of depression on binge drinking was lower for
those with high academic scores than for those with low academic scores. This finding
indicated that the effect of academic score on binge drinking was mitigated by level of
depression among Chinese adolescent females. This finding supported my theoretical model
that hypothesized that intrapersonal domain variables would mitigate the effects of
environmental factors on alcohol use.
Positive statistically significant interactions were also found between depression and
allowance (Std. β = 0.171, p = 0. 0005) and between depression and parental education (Std.
β = 0.261, p <.0001) in their effect on binge drinking. This indicated that the relationship
between binge drinking and depression significantly differed between those with high and
low allowances and between those whose parents had high and low levels of education.
Once again these findings showed that significant interactions were present between

81

intrapersonal psycho-social factors and environmental factors in their association with youth
alcohol use behaviors.
A negative significant interaction was found between depression and parental
drinking (Std. β = -0.136, p = 0.0194) in their effect on binge drinking. This result indicated
that the effect of depression on binge drinking was lower for those with parents who
indicated higher levels of parental drinking. This finding did not support my general
hypothesis that stated that in accordance with the multistage social learning model
adolescents will display heavier alcohol use if they indicated high levels of depression and
had parents who were heavy alcohol users themselves.
Stress
Negative interactions were found between stress and allowance in their effect on
monthly alcohol use (Std. β = -0.107, p = 0.007), and in their effect on binge drinking (Std. β
= -0.145, p = 0.0032). These results indicated that the effect of stress on monthly use and
binge drinking was lower for those with higher allowances. Results in study 1 indicated that
greater disposable income was shown to be associated with binge drinking behaviors. The
results presented in this study showed that stress moderated the effect of allowance on
alcohol use behaviors. The effect of stress on alcohol use behaviors was diminished when
the adolescent had higher disposable income.
Negative interactions were also found between stress and parent education (Std. β =
-0.392, p <.0001) in their effect on binge drinking. This indicated that the effect of stress on
binge drinking was lower for those whose parents had high levels of education. Together
with the results for allowance, this result shows that indicators of higher social standing
interact with psychosocial determinants in their association with AU behaviors among
Chinese adolescents.

82

Hostility
A negative interaction was found between hostility and academic score (Std. β = -
0.095, p = 0.017) in their effect on monthly alcohol use. The negative interaction indicated
that the effect of hostility on monthly alcohol use was lower for those with high academic
scores than with those with low academic scores. This finding confirmed the existence of
significant interactions being present between psychosocial indicators and academic score in
their association with youth alcohol use among Chinese female adolescents. In addition a
positive interaction was found between hostility and academic score (Std. β = 0.123, p =
0.02) in their effect on binge drinking. This indicated that the relationship between binge
drinking and hostility significantly differed between those with high and low academic
scores.

83

Males
Table 12. Environmental and Intrapersonal Variable Interactions, Males

Males N = 6993 Beta Estimates
*

Non Alcohol Users =
2811
Monthly Alcohol
‡
Binge drinking
§

Effect
†
Beta SE P Beta SE P

Parenting
Depression
(log)*Latchkey(log)
ns ns ns ns ns ns

School
Depression
(log)*Academic score
ns ns ns ns ns ns
Depression
Economic
Depression
(log)*Allowance
ns ns ns ns ns ns

Depression (log)*Parental
drinking
ns ns ns ns ns ns

Parental
Depression (log)*Parent
education
ns ns ns ns ns ns
Parenting Stress*Latchkey(log) ns ns ns 0.096 0.041 0.0194
School Stress*Academic score ns ns ns ns ns ns
Stress Economic Stress*Allowance 0.087 0.035 0.014 ns ns ns
Stress*Parental drinking ns ns ns ns ns ns
Parental Stress*Parent education ns ns ns ns ns ns
Parenting Hostility*Latchkey(log) ns ns ns -0.09 0.035 0.0148
School Hostility*Academic score ns ns ns ns ns ns
Hostility Economic Hostility*Allowance ns ns ns ns ns ns
Hostility*Parental drinking ns ns ns ns ns ns
Parental Hostility*Parent education ns ns ns ns ns ns
*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Monthly use: 1 or more drinks of alcohol during the past 30 days.
§:
Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30
days.

Stress
A positive interaction was found between stress and latchkey status (Std. β = 0.096,
p = 0.02) in their effect on binge drinking. This indicated that the relationship between binge
drinking and stress significantly differed between those with high and low time spent home
alone. As was the case in females this result confirms my general theoretical model
hypothesis that postulated the existence of significant interactions between psychosocial
variables and environmental factors in their association with Chinese adolescent AU
behaviors.

84

As was the case in females as well, a positive interaction was found between stress
and allowance (Std. β = 0.087, p = 0.014) in their effect on monthly alcohol use. This
indicates that the relationship between monthly drinking and stress significantly differed
between those with high and low disposable income.
Hostility
A negative interaction was found between hostility and latchkey status (Std. β = -
0.086, p = 0.015) in their effect on binge drinking. This finding indicated that the effect of
hostility on binge drinking was lower for those with higher times spent in self care. This
finding did not support my hypothesis that stated that higher levels of hostility and amount
of time spent in self care would result in higher adolescent alcohol use. Nonetheless, the
presence of a significant interaction does confirm my general hypothesis of intrapersonal
domain factors moderating the effect of environmental factors on adolescent alcohol use.

85

Model 2 Exclusive alcohol use model
This model evaluated the relative odds of our variables of interest predicting a level
of alcohol use between those who endorsed a single level of use and non alcohol users.
Tables 13 and 14 report the results for the exclusive alcohol use model analyses.

Female
Table 13. Environmental and Intrapersonal Variable Interactions, Exclusive Levels of Alcohol
Use, Females

Beta Estimates
*

Non Alcohol Users =
3639
Monthly Alcohol
‡
vs.
Non-use

Binge drinking
§
vs.
Non-use
N = 1140 N = 338
Effect
†
Beta SE P Beta SE P

Parenting
Depression
(log)*Latchkey(log)
0.129 0.046 0.005 ns ns ns

School
Depression
(log)*Academic score
ns ns ns -0.15 0.065 .0202
Depression
Economic
Depression
(log)*Allowance
ns ns ns 0.254 0.059 <.0001

Depression (log)*Parental
drinking
ns ns ns ns ns ns

Parental
Depression (log)*Parent
education
ns ns ns 0.160 0.069 .0201
Parenting Stress*Latchkey(log) -0.10 0.046 0.026 ns ns ns
School Stress*Academic score ns ns ns ns ns ns
Stress Economic Stress*Allowance ns ns ns -0.25 0.058 <.0001
Stress*Parental drinking ns ns ns ns ns ns
Parental Stress*Parent education ns ns ns -0.26 0.071 0.0003
Parenting Hostility*Latchkey(log) ns ns ns ns ns ns
School Hostility*Academic score ns ns ns ns ns ns
Hostility Economic Hostility*Allowance ns ns ns ns ns ns
Hostility*Parental drinking ns ns ns ns ns ns
Parental Hostility*Parent education ns ns ns ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Monthly use: 1 or more drinks of alcohol during the past 30 days but not heavy use.
§:
Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30
days.

86

Depression
A positive interactions was found between depression and latchkey status (Std. β =
0.129, p = 0.005) in their effect on monthly alcohol use. This indicates that the relationship
between monthly alcohol use and depression significantly differed between those with high
and low latchkey status. This finding persisted in both alcohol analyses models indicating a
significant presence of intrapersonal and environmental moderation in youth alcohol use
behaviors.
Positive significant interactions were also found between depression and allowance
(Std. β = 0.254, p <.0001) and between depression and parent education (Std. β = 0.160, p =
0.02) in their effect on binge drinking. This indicated that the relationship between binge
drinking and depression differed between those with high and low allowance, and between
those whose parents had higher levels of education and those whose parents indicated
obtaining lower levels of education. As in the analyses conducted in the general alcohol use
model these findings point to an interaction between psychosocial variables and indicators of
higher social standing.
Stress
A negative interaction was found between stress and latchkey status (Std. β = -0.104,
p = 0.03) in their effect on monthly alcohol use. The effect of stress on monthly AU was
lower for those with higher times spent in self care. Negative interactions were also found
between stress and allowance (Std. β = -0.253, p <.0001) and between stress and parent
education (Std. β = -0.261, p = 0.0003) in their effect on binge drinking. These results
indicated that the effect of stress on binge drinking was lower for those with higher
allowance, and lower for those whose parents indicated completing a higher level of
education. These findings showed that stress acted as a significant moderating agent

87

between various environmental factors and alcohol use among Chinese female adolescents.
This moderating effect persisted in both models of analysis.
Male
Table 14. Environmental and Intrapersonal Variable Interactions, Exclusive Levels of Alcohol
Use, Males

Beta Estimates
*

Non Alcohol Users =
2811
Monthly Alcohol
‡
vs.
Non-use

Binge drinking
§
vs.
Non-use

N = 1292 N = 860
Effect
†
Beta SE P Beta SE P

Parenting
Depression
(log)*Latchkey(log)
ns ns ns ns ns ns

School
Depression
(log)*Academic score
ns ns ns ns ns ns
Depression
Economic
Depression
(log)*Allowance
-0.14 0.052 .006 ns ns ns

Depression
(log)*Parental drinking
ns ns ns -0.1 .052
.013
7

Parental
Depression (log)*Parent
education
ns ns ns ns ns ns
Parenting Stress*Latchkey(log) ns ns ns ns ns ns
School Stress*Academic score ns ns ns ns ns ns
Stress Economic Stress*Allowance 0.191 0.051 .0002 ns ns ns
Stress*Parental drinking ns ns ns ns ns ns
Parental Stress*Parent education ns ns ns ns ns ns
Parenting Hostility*Latchkey(log) ns ns ns ns ns ns
School Hostility*Aca score ns ns ns ns ns ns
Hostility Economic Hostility*Allowance ns ns ns ns ns ns

Hostility*Parental
drinking
ns ns ns ns ns ns

Parental
Hostility*Parent
education
ns ns ns ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Monthly use: 1 or more drinks of alcohol during the past 30 days but not heavy use.
§:
Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30
days.

Depression
A negative interaction was found between depression and allowance (Sβ -0.143 p =
0.006) in their effect on monthly alcohol use. This indicated that the effect of depression on
monthly AU was lower for those who received higher allowance money. As noted

88

previously allowance was found to be significantly associated with alcohol use behaviors in
study 1. This result indicates that depression moderated this relationship among males.
A negative interaction was also found between depression and parental drinking
(Std. β = -0.13, p = 0.01) in their effect on binge drinking. This indicated that the effect of
depression on binge drinking was lower for those whose parents indicated high levels of
alcohol use. This finding did not support my initial hypothesis that expected to find higher
levels of depression and high parental drinking to be associated with higher levels of
adolescent use.
Stress
A significant positive interaction was found between stress and allowance (Std. β =
0.191, p = 0.0002) in their effect on monthly AU. Results showed that the relationship
between monthly alcohol use and stress differed by levels of allowance received. This
finding indicated a persistence of psychosocial variables interacting with allowance in
adolescent alcohol use among both genders and in both levels of analysis.
Model 3: Incremental alcohol use model
Two models were tested, one in which the relative odds of monthly alcohol use
among lifetime alcohol users was assessed and another in which the relative odds of binge
drinking among monthly alcohol users was assessed. Table 15 and 16 report the results for
the relative odds of monthly alcohol use among lifetime alcohol users. Tables 17 and 18
report the results for the relative odds binge drinking among monthly alcohol users.

89

Females
Monthly Alcohol use among Lifetime Alcohol Users Model
Table 15. Relative Odds of Monthly Alcohol Use among Female Lifetime Alcohol
Users, Interaction Analyses

Females N = 3710 Beta Estimates
*

Monthly Alcohol
‡

Effect
†
Beta SE P
Parenting Depression (log)*Latchkey(log) ns ns ns
School Depression (log)*Academic score ns ns ns
Depression Economic Depression (log)*Allowance ns ns ns
Depression (log)*Parental drinking ns ns ns
Parental Depression (log)*Parent education -0.130 0.057 0.023
Parenting Stress*Latchkey(log) ns ns ns
School Stress*Academic score ns ns ns
Stress Economic Stress*Allowance ns ns ns
Stress*Parental drinking ns ns ns
Parental Stress*Parent education ns ns ns
Parenting Hostility*Latchkey(log) ns ns ns
School Hostility*Academic score -0.115 0.054 0.033
Hostility Economic Hostility*Allowance 0.120 0.049 0.015
Hostility*Parental drinking ns ns ns
Parental Hostility*Parent education ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Monthly use: 1 or more drinks of alcohol during the past 30 days but not heavy use.

Depression
A significant negative interaction was identified between depression and parent
education (Std. β = -0.130, p = 0.02) in their effect on monthly AU. This indicated that the
effect of depression on monthly AU was lower for those with parents with high levels of
education. These findings pertained to a restricted level of analysis that was only looking at
the odds of monthly AU among lifetime users.

90

Hostility
A negative interaction was found between hostility and academic score (Std. β = -
0.115, p = 0.03) in their effect on monthly AU. The effect of hostility on monthly use was
lower among those with higher academic scores. A positive interaction was identified
between hostility and allowance (Std. β = 0.120, p = 0.015) in their effect on monthly
alcohol use. This indicates that the relationship between hostility and monthly use differed
by high and low levels of allowance. These findings show that both depression and hostility
acted as significant moderators between environmental factors and youth alcohol use when
lifetime users were at increased odds for monthly AU.
Binge Drinking among Monthly Alcohol Users Model
Table 17. Relative Odds of Binge Drinking Among Female Monthly Alcohol Users,
Interaction Analyses

Females N = 1478 Beta Estimates
*

Binge Drinking
‡

Effect
†
Beta SE P
Parenting Depression (log)*Latchkey(log) ns ns ns
School Depression (log)*Academic score -0.207 0.103 0.0448
Depression Economic Depression (log)*Allowance 0.260 0.101 0.01
Depression (log)*Parental drinking ns ns ns
Parental Depression (log)*Parent education 0.323 0.110 0.0035
Parenting Stress*Latchkey(log) ns ns ns
School Stress*Academic score ns ns ns
Stress Economic Stress*Allowance -0.225 0.110 0.0414
Stress*Parental drinking ns ns ns
Parental Stress*Parent education -0.312 0.121 0.0101
Parenting Hostility*Latchkey(log) ns ns ns
School Hostility*Academic score ns ns ns
Hostility Economic Hostility*Allowance ns ns ns
Hostility*Parental drinking ns ns ns
Parental Hostility*Parent education ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the
past 30 days.

91

Depression
A negative interaction was found between depression and academic score (Std. β = -
0.207, p = 0.04) in their effect on monthly AU. The effect of depression on monthly use was
lower among those with higher academic scores. Positive interactions were found between
depression (Std. β = 0.260, p = 0.01) and allowance (Std. β = 0.323, p = 0.003) and parent
education in their effect on monthly use. This indicated that the relationship between
depression and monthly alcohol use differed by levels of allowance, and by the level of
parental education. These findings showed that depression acted as a significant moderator
between academic score, allowance, parent education and youth alcohol use when monthly
users were at increased odds for binge drinking.
Stress
Negative interactions were identified between stress and allowance (Std. β = -0.225,
p = 0.04) and parental education (Std. β = -0.312, p = 0.01) in their effect on monthly AU.
These results indicated that the effect of stress on monthly AU was lower among those with
high levels of allowance and among those whose parents indicated a high level of education.
These findings indicate that allowance and parental education were moderated by stress and
depression when analysis was restricted to odds of binge drinking among female monthly
users.

92

Male
Monthly Alcohol use among Lifetime Alcohol Users Model
Table 16. Relative Odds of Monthly Alcohol Use among Male Lifetime Alcohol Users,
Interaction Analyses

Males N = 4138 Beta Estimates
*

Monthly Alcohol
‡

Effect
†
Beta SE P
Parenting Depression (log)*Latchkey(log) ns ns ns
School Depression (log)*Academic score ns ns ns
Depression Economic Depression (log)*Allowance ns ns ns
Depression (log)*Parental drinking ns ns ns
Parental Depression (log)*Parent education ns ns ns
Parenting Stress*Latchkey(log) ns ns ns
School Stress*Academic score ns ns ns
Stress Economic Stress*Allowance 0.106 0.052 0.0446
Stress*Parental drinking ns ns ns
Parental Stress*Parent education ns ns ns
Parenting Hostility*Latchkey(log) ns ns ns
Hostility*Parenting style ns ns ns
School Hostility*Academic score ns ns ns
Hostility Economic Hostility*Allowance ns ns ns
Hostility*Parental drinking ns ns ns
Parental Hostility*Parent education ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Monthly use: 1 or more drinks of alcohol during the past 30 days but not heavy use.

Stress
A positive interaction was found between stress and allowance (Std. β = 0.106, p =
0.04) in their effect on binge drinking. The effect of stress on binge drinking differed
between those who receive low allowance and those who received high allowances.
Significant interactions between stress and allowance were shown to be persistent across all
levels of alcohol analysis among males. Indeed the interaction between these variables was
shown to be present when analyses were restricted to odds of monthly use among lifetime
male alcohol users.

93

Binge Drinking among Monthly Alcohol Users Model
Table 18. Relative Odds of Binge Drinking Among Male Monthly Alcohol Users,
Interaction Analyses

Males N = 2152 Beta Estimates
*

Binge Drinking
‡

Effect
†
Beta SE P
Parenting Depression (log)*Latchkey(log) ns ns ns
School Depression (log)*Academic score ns ns ns
Depression Economic Depression (log)*Allowance ns ns ns
Depression (log)*Parental drinking ns ns ns
Parental Depression (log)*Parent education ns ns ns
Parenting Stress*Latchkey(log) ns ns ns
School Stress*Academic score ns ns ns
Stress Economic Stress*Allowance -0.174 0.070 0.013
Stress*Parental drinking ns ns ns
Parental Stress*Parent education ns ns ns
Parenting Hostility*Latchkey(log) ns ns ns
School Hostility*Academic score ns ns ns
Hostility Economic Hostility*Allowance ns ns ns
Hostility*Parental drinking ns ns ns
Parental Hostility*Parent education ns ns ns

*
: Adjusted for age, parent drinking, income, parent education, city, grade, district rank.
†
: Standardized scores used for predictors of interest.
‡:
Binge Drinking: 5 or more drinks of alcohol in a row within a couple of hours during the past 30
days.

Stress
Once again the significant interaction between stress and allowance was found when
analyses were restricted to odds of binge drinking among male alcohol users. A significant
negative interaction was found between stress and allowance (Std. β = -0.174, p = 0.01) in
their effect on binge drinking. The effect of stress on binge drinking was lower among those
who receive higher allowances.
Discussion
Results from this study showed a number of significant interactions between
intrapersonal and environmental domain factors. Findings indicated that the associations

94

between the selected intrapersonal variables and both levels of alcohol use differed mainly
according to academic score, allowance, latchkey status, and parental education.
Results indicated that the effect of intrapersonal factors on alcohol use behaviors
differed by level of academic achievement among both genders. These effects indicated that
analyses of the role of intrapersonal variables on alcohol use behaviors that include academic
score variables in the model should consider including stratified levels of academic scores in
order to conduct further analyses within each academic score subgroup. This information
could serve to inform prevention efforts so that academic components can be included as
part of efforts to delay onset and curb progression to heavier forms of use. Cultural
investigations in the US and in China have stressed the importance of academic goals in
Chinese populations (Chao 1994; Chao 1995; Chao and Sue 1996). The results presented
here should be considered of critical importance for alcohol use prevention efforts in this
population.
Results also indicated that the relationships between selected intrapersonal variables
and both levels of alcohol use differed by levels of allowance. Once again these findings
indicated that further analysis should be conducted when intrapersonal factors and
disposable income are being investigated for their role on alcohol use behaviors. These
analyses should include stratified subgroups of allowance in order to better comprehend the
influence that this variable has on youth alcohol use behaviors. Further investigations
should place emphasis on what Chinese adolescents do with their disposable income. The
information presented here can certainly be included as part of education efforts that teach
parents and caregivers the ways in which disposable income may lead to unhealthy behavior
choices, especially in cases where their youths have co-morbid or pre-existing psychosocial
disorders.

95

Findings in this study also indicated that the associations between alcohol use and
psychosocial variables varied according to latchkey status. Further analyses are required
which can clarify the effects that intrapersonal factors have on alcohol use behaviors
according to different amounts of time spent in self care. The moderation effect that exists
between the selected psychosocial variables and this parenting practice represents a novel
contribution to the field of adolescent alcohol use. Many investigations have demonstrated
associations between intrapersonal and environmental factors in their association with youth
alcohol use behaviors. However this investigation is among the first to demonstrate the
moderating effects of psychosocial attributes on the relationship between latchkey status and
Chinese adolescent AU behaviors.
With regards to parental education results showed that that the relationship between
alcohol use and psychosocial attributes differed significantly between those whose parents
indicated high levels of education and those whose parents indicated completing low levels
of education in both models of analysis. A greater emphasis must be put into clarifying the
role of parental demographic characteristics on youth alcohol use. Indeed, findings from this
study and the previous one have shown that these variables do associate with youth alcohol
use behaviors. The results presented here should serve as basis for prevention efforts that
educate parents and caregivers as to their role in facilitating youth alcohol use onset and use
disorders like binge drinking.
The findings presented in this study support my general theoretical model that posits
the moderating effects of intrapersonal domain factors on the relationship between
environmental factors and youth alcohol use. The interaction effects presented in this study
should serve as a first step in uncovering the complex relationships that exist between the
various determinants that influence Chinese adolescent AU behaviors. This investigation

96

has served to further our understanding of the nature of the associations that exist among the
different variables that concurrently affect youth AU behaviors. An important contribution
of this study to the field of youth alcohol youth is that we have shown the presence of
moderating effects across two levels of alcohol use and three distinct models of analysis.
Further investigations are required which can replicate these findings among different study
populations. The information garnered from these types of studies can serve to inform
prevention efforts that take into account the complicated interactions that may exist between
AU risk factors. Prevention scientists must be made aware that intrapersonal and
environmental AU influences most likely will affect one another in their relationship to AU
behavior outcomes.

97

CHAPTER SIX: A FOLLOW UP LATENT GROWTH CURVE ANALYSIS OF
INTRAPERSONAL AND ENVIRONMENTAL DETERMINANTS OF ALCOHOL
USE IN CHINESE YOUTH
As we have reviewed, some of the more important adolescent alcohol use
determinants identified in the literature pertain to both the intrapersonal and environmental
domains of youths’ lives. Common risk factors associated with alcohol use initiation include
parental drinking and substance use, peer alcohol use, parental monitoring of child
behaviors, time spent in self-care, tolerant attitudes toward deviance or approval of alcohol
and drug use, affective disorders (depression, anxiety, stress), engaging in other problem
behaviors, and low bonding to conventional institutions (Hawkins and Weis 1985; Hawkins,
Catalano et al. 1992).
Many effective substance use prevention interventions have focused on some of
these determinants to prevent or delay the onset of substance use during early adolescence.
Early onset has been shown to increase the risk for use in late adolescence, and to increase
the risk for abuse and dependence later in life (Mayhew, Flay et al. 2000; Donovan and . :
2004; Glantz and Mandel 2005; Sher, Grekin et al. 2005). A number of successful
interventions have shown effects in preventing youth alcohol use initiation and in preventing
abuse and alcohol use disorders in late adolescence et al. 2004; Glantz and Mandel 2005;
Sher, Grekin et al. 2005).
While studies identifying correlates of alcohol use serve the important function of
furthering our understanding of alcohol use behaviors, a number of investigators have
pointed out that there is a great need in the field to test more complex models of alcohol use
that analyze predictors for alcohol use transition behaviors (Donovan, Jessor et al. 1983;
Donovan and Jessor 1985; Donovan et al. 2004). We know that for an individual to develop

98

alcohol use dependence and subsequent disorders he or she must first pass through a stage of
alcohol use experimentation. It is during this stage that particular risk factors are implicated
in aiding transition from experimental consumption behaviors to riskier forms of use and
subsequent use disorders. In order to identify the determinants that are implicated in the
transition to riskier alcohol use behaviors we must be able to identify risk factors that predate
the onset of abuse and disorder. Risk factors for alcohol abuse and for the development of
use disorders need to be shown to be statistically significantly associated to problematic use
behaviors and their needs to be evidence that these factors were present before the
development of such alcohol use behavior problems.
The determinants that are commonly associated with alcohol use experimentation
are rarely tested for their involvement in adolescent alcohol use transition behaviors.
Adolescent substance use investigators have pointed to the fact that many etiological studies
conducted on adolescent substance use tend to focus only on psychosocial variables that
predict initiation or that distinguish users from non-users (Mayhew, Flay et al. 2000;
Donovan and . : 2004; Glantz and Mandel 2005; Sher, Grekin et al. 2005).
These study outcomes may confound the process of onset and of escalation of use,
and may fail to take into account that there might be different predictors for SU onset and for
SU transition behaviors (Maggs, Schulenberg et al. 1997). Only 6 of a total 54 recent
empirical studies that identified adolescent alcohol use determinants analyzed longitudinal
AU transition behaviors in youth populations [Table A].
Research is needed which can allow researchers to identify the determinants that
lead to alcohol use onset and which lead to progression to heavier alcohol use. Efforts must
identify variables that serve as protective factors against AU problem behaviors.

99

In the previous two cross-sectional investigations I have attempted to learn the
effects that particular determinants of AU behaviors have on adolescent alcohol use by
comparing these effects on different levels of alcohol use and non-use. What these cross-
sectional investigations have not been able to determine is whether changes in these
determinants or in alcohol use at baseline change the level of alcohol use at follow up. In
order to obtain this type of information we can utilize latent growth curve analysis, which is
a method to study just such changes.
Considering the findings from the previous two studies in this dissertation proposal
this study investigated whether changes in selected alcohol use determinants influenced
changes in level of alcohol use at follow up. Additionally this study investigated if selected
variables moderated the impact of other variables in their relationship to alcohol use
progression.
This study examined changes in levels of alcohol use over two to three waves of
data utilizing growth curve analysis. This study investigated if changes in selected
determinants increased or reduced levels of alcohol use at follow up. This study was based
on sequential cohort data from the CSCS to model the trajectory of youth level of alcohol
use over 2 to 3 waves of follow up data. Three waves of data were analyzed for students in
grades 7 and 8 at baseline, and two waves of data were analyzed for students who were
enrolled in grades 8 and 11 at baseline. Taking into account the results form both cross-
sectional studies in this dissertation I estimated the potential influence that the intrapersonal
factors: depression, stress, hostility; together with 30 day smoking, junk food use, and
academic score had on alcohol use change during the stated time periods. These analyses
were conducted on two levels of alcohol use: monthly alcohol use and on binge drinking.

100

The variables 30 day smoking, junk food use, and academic score were selected
from the pool of available variables due to the robustness of their associations in the
previous two cross-sectional studies. In addition to the strength of their associations with all
levels of alcohol use in the previous studies these variables were selected because they
allowed me to formulate theoretical considerations which contributed to the field of youth
alcohol research. The incremental associations found between junk food use and increasing
levels of alcohol use in study one represented novel findings to the field of alcohol use. This
variable required more detailed investigation which can help determine its role in follow up
alcohol use behaviors. Likewise academic score proved to be one of the strongest protective
factors identified in study one, and also proved to be a significant moderated by the selected
intrapersonal domain factors in its association with two levels of alcohol use. 30 day
smoking represented an important variable to investigate for its possible role in alcohol use
changes given the robustness it showed in associations with all levels of alcohol use. Studies
in the field of substance use have reported strong findings for the presence of a smoking and
alcohol use co-morbidity (Jackson, Henriksen et al. 1997; Jackson, Sher et al. 2005).
Consequently it was important to study this phenomenon in an adolescent population that
has been found to have high use rates for both substances.
In addition, two separate moderation analyses were conducted. The first moderation
analyses analyzed interactions between each intrapersonal factor and 30 day smoking, junk
food use, and academic score in their relationship to two alcohol use levels (monthly use,
and binge drinking). The second moderation analyses analyzed interactions between each
intrapersonal factor and three parenting variables measured at baseline: parent monitoring,
latchkey status, and parenting style. I chose to analyze interactions between each

101

intrapersonal factor and these parenting variables because these variables are consistently
implicated with adolescent AU behaviors. I focused my analyses on baseline values of these
variables because they were not assessed for the full 2 wave follow up.
Hypotheses
Consistent with the third stage of the multistage social learning model (Simons,
Conger et al. 1988) which posits that adolescent alcohol use experimentation may transition
to heavy alcohol use behaviors if the adolescent indicates emotional distress, and poor
coping skills it was expected that:
(1) Adolescents with high levels of depression at baseline would be positively
related to increases in alcohol use and binge drinking behaviors at follow-up.
(2) Adolescents with high levels of hostility at baseline would be positively related
to increases in alcohol and binge drinking behaviors use at follow-up.
(3) Adolescents with high levels of stress at baseline would be positively related to
increasing levels of alcohol use and binge drinking behaviors at follow-up.
(4) High levels of funk food use and 30 day smoking would be positively related to
increases in alcohol and binge drinking behaviors use at follow-up.
(5) High academic scores would influence a decrease in alcohol use at follow up.
(6) High parental monitoring would prove to be a significant moderator for the
relationship between the selected intrapersonal variables and alcohol use.
(6) High latchkey status would prove to be a significant moderator for the
relationship between the selected intrapersonal variables and alcohol use.
(6) Academic score would prove to be a significant moderator for the relationship
between the selected intrapersonal variables and alcohol use.

102

Data Analysis
Growth curve analyses were used to estimate the effects of our intrapersonal and
environmental variables on levels of alcohol use over time. Analyses were performed with
the Mplus statistical package (v5 Muthen & Muthen). Growth curve analysis is a technique
that can be used to model longitudinal change in repeated observations of a dependent
variable (Duncan, Duncan et al. 1999) The growth curve analysis approach takes advantage
of both structural equation modeling, which incorporates latent variables, and hierarchical
linear modeling which allows random coefficients across individual developmental
trajectories (Simons-Morton, Chen et al. 2004).
Growth trajectory is often specified as a linear function of time, in which case it contains two
important unknown individual growth parameters: an intercept and a slope that determine
the shape of individual true growth over time (Lenzenweger, Johnson et al. 2004). The
intercept parameter represents the net elevation of the trajectory over time (the true mean of
alcohol use level at the onset of the study. The slope parameter represents the rate of change
over time. Three waves of data were analyzed for students in grades 7 and 8 at baseline, and
two waves of data were analyzed for students who were enrolled in grades 8 and 11 at
baseline.
The analysis will be conducted in two steps:
First, an unconditional growth curve models (without covariates) will be estimated
to determine the shape of the developmental trajectory of adolescent drinking stage
progression over time (normative trajectory). In fitting the unconditional growth
curve models, a significant variance in intercept will reveal substantial individual
differences in drinking status at baseline. Significant variation in latent growth

103

factors (such as the slope in a linear model) will indicate individual differences in
the probability of progressing in drinking level over time.
Likewise I will estimate unconditional growth curve models separately for each risk
factor to determine the shape of the developmental trajectories for these variables,
and to determine if significant individual differences in change occur over time for
these variables.
Second, conditional growth curve models will be used to evaluate the relationships
of the study covariates and the intrapersonal and environmental variables with
differences in initial drinking level and progression over time. A significant path
coefficient from a covariate leading to the latent intercept will indicate that this
covariate is associated with individual differences in drinking level at baseline. A
significant covariate leading to the growth factors would reveal an association
between this covariate and individual differences in the probability of drinking level
progression over time. Possible confounders adjusted for in the regressions will
included age, grade, puberty, city, type and academic level of school attended (poor,
medium, good, professional/vocational), economic level of municipal district (poor,
medium, good), parental education level, parental income level, and parental heavy
alcohol use. Analyses will be conducted separately by gender in order to assess
gender differences that may exist in these associations. Model fit will be assessed
using the comparative fit index (CFI).

104

Results
Female, Monthly Alcohol Use
Stress, 30 day smoking
Figure 2. Longitudinal Analysis of Stress, Monthly Drinking, 30 Day Smoking, Females
Only

Adjusted for: age, school rank, district rank, city

Figure 2 shows that initial high levels of stress were correlated to initial high levels
of 30 day smoking (Std. β = 0.377, p < 0.001). Results also indicated that initial levels of
stress were correlated with initial monthly drinking (Std. β = 0.160, p < 0.001). This finding
supports the association between 30 day smoking and monthly AU in study one. Regarding

Month
AU
Wave 1
Month
AU
Wave 2

Month
AU
Wave 3

Stress
Wave 3

Stress
Wave 2

Stress
Wave 1

SMK
Wave 3

SMK
Wave 1

SMK
Wave 2
30 Day
SMK
Slope
Initial 30
Day SMK

Initial
Stress

Stress
Slope

Monthly
AU Slope
Initial
Monthly
AU
2
1
1 1
1
2
1
1
1
1
2
1
1 1 1
0.377 (0.000)
0.160 0.000) -0.002
0.025 (0.024)

105

changes of initial values leading to changes in slopes we see that initial levels of
stress influenced change of stress at follow up (Std. β = 0.025, p = 0.024). This finding
indicates that initial high stress influenced a positive change in follow up stress when 30 day
smoking was in the model. Results showed that initial high levels of monthly drinking led to
decrease in stress at follow up (Std. β = -0.002, p = 0.025). This indicates that monthly
alcohol consumption reduced the experience of high stress at follow up in adolescent
females. This would seem to show that Chinese adolescent females may be utilizing
monthly AU as a method to reduce stress. The comparative fit index for this model (CFI =
.972) indicated satisfactory model fit.
Junk food use
Figure 3. Longitudinal Analysis of Stress, Monthly Drinking, Junk Food Use, Females Only

Adjusted for: age, school rank, district rank, city

Initial Junk
Food Use

Junk Food
Use Slope

Initial
Stress
Stress
Slope
Stress
Junk Food Use

Initial
Monthly
AU
Monthly
AU Slope

Monthly Drinking
0.162 (0.000)
0.147 (0.000)
0.320 (0.000)
0.496 (0.000)
-0.026 (0.029)

106

For clarity purposes from here on out only the model factors will be displayed in the
figures. Figure 3 indicates that initial levels of stress were correlated to initial levels of junk
food use (Std. β = 0.320, p < 0.001) and to initial levels of monthly alcohol use (Std. β =
0.162, p < 0.001). As reported in study one, initial levels of monthly drinking were
correlated to initial levels of junk food use (Std. β = 0.147, p < 0.001). These findings show
that follow up estimates of junk food use were also found to be correlated to follow up
estimates of monthly AU (Std. β = 0.496, p < 0.001). Interestingly, results for this model
indicated that initial levels of Monthly AU led to a decrease in follow up of monthly AU
(Std. β = -0.026, p = 0.029). This finding contradicts those reported in Western populations
that indicate that early adolescent alcohol use tends to be associated with use later in
adolescence. The correlations that are present in this model between initial levels of stress,
junk food use, and monthly AU, and between slope values of junk food use and monthly AU
indicate that this model represents a co-morbidity model of monthly alcohol use. The
comparative fit index for this model (CFI = .973) indicated satisfactory model fit.

107

Academic Score
Figure 4. Longitudinal Analysis of Stress, Monthly Drinking, Academic Score, Females
Only

Adjusted for: age, school rank, district rank, city

Results (see figure 4) indicated an inverse correlation between high initial stress and
initial academic score values (Std. β = -0.109, p < 0.001). This indicates that initial high
stress was correlated with initial low academic scores. Initial academic scores also showed
an inverse correlation with monthly AU (Std. β = -0.090, p < 0.001) which indicated that
lower academic scores were correlated with higher levels of monthly drinking. These
findings are similar to those of study one which showed high academic scores to be
protective for alcohol use. High levels of initial stress were correlated to high initial
monthly alcohol use (Std. β = 0.162, p < 0.001). Results indicated that initial stress

Initial
Academic
Score
Academic
Score Slope

Initial
Stress
Stress
Slope
Stress
Academic Score

Initial
Monthly
AU
Monthly
AU Slope

Monthly Drinking
0.162 (0.000)
-0.090 (0.000)
-0.109 (0.000)
0.013 (0.000)
0.023 (0.015)
0.606 (0.023)

108

influenced change in academic score at follow up (Std. β = 0.606, p = 0.023). This indicated
that high initial stress was found to positively influence change in academic scores at follow
up. In addition initial academic score was found to influence change of academic score at
follow up (Std. β = 0.023, p = 0.015). Both these findings may point to the importance given
to high academic achievement among Chinese adolescent females. It is then noteworthy to
see that initial high academic scores were found to influence change in monthly drinking at
follow up (Std. β = 0.013, p < 0.001). This finding indicates that initial high academic
scores positively related to monthly alcohol use change at follow up. The relationships
between high stress, academic achievement and positive change in monthly AU are
noteworthy in this model. The comparative fit index for this model (CFI = .971) indicated
satisfactory model fit.

109

Hostility
Figure 5. Longitudinal Analysis of Hostility, Monthly Drinking, Academic Score, Females Only

Adjusted for: age, school rank, district rank, city

Academic Score
Similar to results in the previous model which showed a psychosocial variable to
influence change in academic achievement at follow up, results for this model indicated
(figure 5) that high initial hostility was found to influence academic score change at follow
up (Std. β = 0.773, p < 0.001). On the other hand, initial high levels of monthly alcohol use
were found to decrease academic scores at follow up when hostility was in the model (Std. β
= -0.107, p = 0.005). As was the case when stress was analyzed in the model, initial high
academic scores were found to positively influence change in academic scores at follow up
(Std. β = 0.159, p = 0.007). Initial high monthly AU was found to decrease monthly AU at
Initial
Academic
Score
Academic
Score Slope

Initial
Hostility
Hostility
Slope
Hostility
Academic Score

Initial
Monthly
AU
Monthly
AU Slope
Monthly Drinking
0.170 (0.000)
-0.110 (0.000)
-0.104 (0.000)
0.159 (0.007)
-0.069 (0.001)
-0.107 (0.005)
0.773 (0.013)

110

follow up as well (Std. β =- 0.069, p = 0.001). Regarding model correlations initial hostility
was found to be inversely correlated with initial academic score (Std. β = -0.104, p < 0.001).
This indicates that high initial hostility was correlated with low initial academic scores.
Initial academic scores were also found to be inversely correlated with initial monthly AU
(Std. β = -0.110, p < 0.001). This indicated that low initial academic scores were associated
with initial high monthly alcohol use. This finding corresponds with those presented in
study one which showed high academic scores to be protective for monthly AU. Initial
hostility was found to be correlated with initial monthly AU (Std. β = 0.170, p < 0.001). The
comparative fit index for this model (CFI = .971) indicated satisfactory model fit.
Male
Binge Drinking, Depression, 30 day smoking
Figure 6. Longitudinal Analysis of Depression, Binge Drinking, 30 day Smoking, Males Only

Adjusted for: age, school rank, district rank, city
Initial 30 Day
Smoking

30 Day
Smoking Slope

Initial
Depression

Depression
Slope

30 day Smoking

Initial
Binge AU
Binge AU
Slope
Binge Drinking
0.184 (0.000)
0.581 (0.000)
0.650 (0.000)
-0.364 (0.010)
(0.000)
Depression
0.361 (0.000)

111

Figure 6 shows that high levels of initial depression influenced a decrease in binge
drinking when 30 day smoking was present (Std. β = -0.364, p = 0.010). This finding
contradicts those reported in Western youth populations that have found initial levels of
depression to be associated with heavy use later in adolescence. Initial depression was found
to be correlated with both initial smoking (Std. β = 0.581, p < 0.001) and as reported in study
one with initial binge drinking (Std. β = 0.361, p < 0.001). Also similar to findings in study
one, initial binge drinking was correlated with initial levels of 30 day smoking (Std. β =
0.184, p < 0.001). Results show that in addition to the initial values being correlated
between these two variables, their slopes indicated a strong significant correlation as well
(Std. β = 0.6500, p < 0.001). As was the case in female analysis these results are indicative
of co-morbidity between 30 day smoking and binge drinking. The comparative fit index for
this model (CFI = .967) indicated satisfactory model fit.

112

Junk food use
Figure 7. Longitudinal Analysis of Depression, Binge Drinking, Junk Food Use, Males Only

Adjusted for: age, school rank, district rank, city

Initial high depression was correlated with initial high junk food use (Std. β = 0.269,
p < 0.001) and with high initial binge drinking (Std. β = 0.366, p < 0.001). High initial junk
food use was correlated with high binge drinking as well (Std. β = 0.167, p < 0.001). This
corresponds with findings in study one which showed high junk food consumption to be
associated with binge drinking behaviors. As in the previous model Figure 7 shows that
high levels of initial depression influenced a decrease in binge drinking when junk food use
was present (Std. β = -0.480, p < 0.001). As reported in studies in Western populations

Initial Junk
Food Use
Junk Food
Use Slope

Initial
Depression
Depression
Slope

Junk Food Use

Initial
Binge AU
Binge AU
Slope
Binge Drinking
0.366 (0.000)
0.167 (0.000)
0.269 (0.000)
-0.480 (0.000)
Depression
0.057 (0.029)

113

initial binge drinking was found to influence change in binge drinking at follow up (Std. β =
0.057, p = 0.029). The comparative fit index for this model (CFI = .950) indicated
satisfactory model fit.
Academic Score
Figure 8. Longitudinal Analysis of Depression, Binge Drinking, Academic Score, Males Only

Adjusted for: age, school rank, district rank, city

Similar to results found among females which showed psychosocial attributes to
positively influence academic scores, Figure 8 shows that high levels of initial depression
were found to positively influence change in academic score at follow up (Std. β = 0.930, p
= 0.021). As in the previous models, high levels of initial depression were also found to
influence a decrease in binge drinking when academic score was present (Std. β = -0.431, p
< 0.001). Initial high levels of depression were inversely correlated with initial academic
scores (Std. β = -0.155, p < 0.001), and positively correlated with initial binge drinking (Std.
Initial
Academic
Score
Academic
Score Slope
Initial
Depression
Depression
Slope

Academic Score

Initial
Binge AU
Binge AU
Slope
Binge Drinking
-0.051 (0.030)
-0.155 (0.000)
-0.398 (0.000)
-0.431 (0.001)
0.930 (0.021)
0.170 (0.014)
Depression
0.364 (0.000)

114

β = 0.364, p < 0.001). Initial academic scores were inversely correlated with binge drinking
as well (Std. β = -0.051, p = 0.030). These results indicate that initial low academic scores
were correlated with high initial levels of depression and binge drinking. Slopes between
academic scores and binge drinking were also inversely correlated (Std. β = -0.398, p =
0.030). The comparative fit index for this model (CFI = .958) indicated satisfactory model
fit.
Stress
Figure 9. Longitudinal Analysis of Stress, Binge Drinking, 30 day Smoking, Males Only
Adjusted for: age, school rank, district rank, city
30 day smoking
Results (see figure 9) indicated a correlation between initial stress and initial 30 day
smoking (Std. β = 0.584, p < 0.001) and between initial stress and initial binge drinking (Std.
β = 0.160, p < 0.001). Initial 30 day smoking was correlated with initial binge drinking (Std.
Initial 30 Day
Smoking

30 Day
Smoking Slope

Initial
Stress
Stress
Slope

30 day Smoking

Initial
Binge AU
Binge
AU Slope
Binge Drinking
0.160 (0.000)
0.075 (0.006)
0.584 (0.000)
0.713 (0.000)
Stress

115

β = 0.075, p = 0.006); as were the slopes between these to variables (Std. β = 0.713, p <
0.001). As was the case in the female analyses, the correlations that were present in this
model between initial levels of stress, 30 day smoking, and binge drinking, and between
slope values of 30 day smoking and binge drinking indicated that this model represented a
co-morbidity model of binge drinking and smoking behavior. The comparative fit index for
this model (CFI = .974) indicated satisfactory model fit.
Junk food use
Figure 10. Longitudinal Analysis of Stress, Binge Drinking, Junk Food Use, Males Only

Adjusted for: age, school rank, district rank, city

Figure 10 shows that high levels of initial junk food were found to decrease binge
drinking at follow up (Std. β = -0.031, p = 0.019). This finding does not support my
hypotheses which stated that high levels of junk food use would positively influence change
in binge drinking at follow up. Further analyses are required which can explore why it is
Initial Junk
Food Use

Junk Food
Use Slope

Initial
Stress
Stress
Slope

Junk Food Use

Initial
Binge AU
Binge
AU Slope
Binge Drinking
0.159 (0.000)
0.083 (0.003)
0.271 (0.000)
-0.031 (0.019)
-0.273 (0.041)
-0.061 (0.050)
Stress
-0.036 (0.024)

116

that two variables whose high levels were significantly associated at baseline actually show
an inverse relationship at follow up. Initial high levels of junk food use were also found to
decrease high levels of junk food use at follow up (Std. β = -0.061, p = 0.050). Thus, high
junk food use at baseline was found to decrease both binge drinking and junk food use at
follow up. As in previous models initial high levels of stress were found to decrease binge
drinking at follow up as well (Std. β = -0.273, p = 0.041). In accordance with results from
study one, initial junk food use was correlated with initial binge drinking (Std. β = 0.083, p =
0.003). The comparative fit index for this model (CFI = .971) indicated satisfactory model
fit.
Academic Score
Figure 11. Longitudinal Analysis of Stress, Binge Drinking, Academic Score, Males Only
Adjusted for: age, school rank, district rank, city
Initial
Academic
Score
Academic
Score Slope

Initial
Stress
Stress
Slope

Academic Score

Initial
Binge AU
Binge
AU Slope
Binge Drinking
-0.082 (0.000)
-0.154 (0.000)
-0.293 (0.027)
0.759 (0.020)
0.141 (0.013)
Stress
0.156 (0.000)

117

Again, as previous models have shown, an initial psychosocial indicator was found
to decrease binge drinking at follow up. Figure 11 shows initial high stress to decrease binge
drinking at follow up (Std. β = -0.293, p = 0.027). Likewise initial high stress positively
influenced change in academic score at follow up (Std. β = 0.759, p = 0.020). As was he case
in the female analysis initial academic score was related to change in academic score at
follow up (Std. β = 0.141, p = 0.013). Initial stress was found to be inversely correlated with
initial academic score (Std. β = -0.154, p < 0.001), and positively correlated with initial
binge drinking (Std. β = 0.156, p <0.001). In accordance with findings from study one initial
academic score was inversely related to binge drinking (Std. β = -0.082, p < 0.001).
Correlation results for academic score showed that initial low academic scores were
correlated with high initial stress and high initial binge drinking. The comparative fit index
for this model (CFI = .973) indicated satisfactory model fit.

118

Hostility
Academic Score
Figure 12. Longitudinal Analysis of Hostility, Binge Drinking, Academic Score, Males Only
Adjusted for: age, school rank, district rank, city
Similar to the results presented in the female analysis where stress and academic
performance were analyzed, results indicated that high academic scores positively influenced
change in binge drinking at follow up (Std. β = 0.003, p = 0.021). High levels of hostility
were found to positively influence change in binge drinking (Std. β = 0.150, p = 0.038) and
academic score (Std. β = 0.989, p < 0.006) at follow up as well. In contrast to findings
reported among Western youth, initial binge drinking decreased binge drinking at follow up
(Std. β = -0.095, p < 0.001). Results showed initial hostility to be correlated with binge

Initial
Academic
Score
Academic
Score Slope

Initial
Hostility
Hostility
Slope

Academic Score

Initial
Binge AU
Binge
AU Slope
Binge Drinking
0.231 (0.000)
-0.151 (0.000)
0.003 (0.021)
0.150 (0.038)
0.989 (0.006)
0.178 (0.003)
Hostility
-0.095 (0.000)

119

drinking (Std. β = 0.231, p < 0.001), and inversely correlated with academic score
(Std. β = -0.151, p = 0.003). This indicated that high initial levels of hostility were correlated
with low initial academic scores. The comparative fit index for this model (CFI = .964)
indicated a satisfactory model fit.
Moderation Analysis
No significant interactions were found between any intrapersonal factor and 30 day
smoking, junk food use, and academic score in their relationship to the two alcohol use
levels (monthly use, and binge drinking). No significant interactions were found between
any intrapersonal factor and the three parenting variables measured at baseline: parent
monitoring, latchkey status, and parenting style.
Discussion
Latent growth curve analysis was utilized to analyze if selected alcohol use
determinants influenced changes in level of youth of alcohol use over a 2 wave follow up.
Results presented in this study indicated causal models to explain the follow up alcohol use
behaviors of our Chinese youth study population.
In females results indicated that monthly alcohol use reduced the experience of high
stress at follow up when 30 day smoking was analyzed in the model. It is worth noting that
initial levels of stress were observed to be correlated with initial levels of 30 day smoking
and with initial levels of monthly drinking in this same model. This is a noteworthy finding
that requires more attention. Further investigations are necessary which can clarify the role
of monthly alcohol use as a stress regulator among adolescent females. If monthly drinking
is indeed acting as a stress regulator in females this represents a critical piece of information
for intervention development. The available presence of alternative stress reduction

120

techniques for youth may help delay alcohol use onset in those looking for a quick ways to
reduce stress.
Interestingly, in males initial high levels of stress were found to decrease binge
drinking when junk food use was analyzed in the same model. This same finding was
present when academic score was analyzed in the same model. These findings contradict my
expected findings that hypothesized that higher levels of stress would be associated with
positive changes in alcohol use behaviors at follow up. Studies suggest that those
individuals whose coping styles are primarily limited to alcohol use as a means of combating
stress comprise a potential vulnerable population to alcohol use onset and misuse(Abrams
and Niaura 1987; Wills and Hirky 1996). Indeed, it has been suggested that alcohol use may
develop in part due to the fact that an individual’s coping capacity or skills are not sufficient
to deal with situational demands (Abrams and Niaura 1987). It may be the case that Chinese
adolescent males are finding a way to cope with high stress and that these coping
mechanisms are delaying onset of later alcohol use behaviors. Further investigations are
required which can help determine if cultural specific factors are helping Chinese male
adolescents cope with stress in efficient manners that circumvent unhealthy stress reduction
choices.
Similar to this finding in males, results showed that high levels of initial depression
influenced a decrease in binge drinking when 30 day smoking was present. This finding
seems to contradict the findings presented in study 1 that indicated increasing levels of
depression to be associated with incremental levels of alcohol use. However this same
causal relationship was present when junk food use and academic scores were analyzed in
the same model. In both cases initial levels of depression were strongly correlated with

121

initial levels of binge drinking and with initial levels of 30 day smoking and junk food use
respectively. The slopes of 30 day smoking and binge drinking were also found to be
correlated which may indicate co-morbidity between these to variables in Chinese youth.
Further investigations are required which can help determine the reasons behind the fact that
initial high levels of depression seem to be influencing decreases in binge drinking in males
at follow up. This finding is counterintuitive with findings in the literature that strongly
associate depression with later alcohol use. More detailed investigations which explore both
cultural and biological determinants of adolescent Chinese youth are necessary in order to
gain a more complete picture of what may be occurring in this case. We know that some
Asian populations are less likely to become alcohol dependent compared to other ethnic
groups given particular physiological characteristics such as deficiencies in aldehyde
dehydrogenase (ALDH2) (Lin and Cheng 2002) (Hao, Chen et al. 2005). It is estimated that
approximately half of the Chinese population exhibit a deficiency in aldehyde
dehydrogenase (ALDH2), which is responsible for metabolizing acetaldehyde into acetic
acid, a component of the alcohol-metabolizing human biochemical pathway (Lin and Cheng
2002). Polymorphisms of ALDH2 genes have been found to modify alcohol-drinking
behaviors and possible risks associated with alcoholism (Lin and Cheng 2002). Further
investigations are required which can include depression as part of these analyses.
Perhaps the more interesting findings of this study however, related to the influence
of academic scores on follow up alcohol use behaviors. Initial high academic scores were
found to influence change in monthly drinking in females when stress was present in the
model. In addition, initial high academic scores were found to influence change in binge
drinking among males when hostility was in the model. These findings indicate that initial

122

high academic scores related positively to changes in monthly alcohol use and binge
drinking behaviors at follow up. Furthermore in females high initial stress was found to
positively influence change in high academic scores at follow up. In males high initial
depression and high initial hostility were found to positively influence change in academic
score at follow up. In both cases we see that initial high levels of psychosocial factors
positively influenced academic scores at follow up in both genders.
These findings are unique in that in the West investigations have shown that
increased substance use may occur among those with failing or substandard academic
achievements (Kumpulainen 2002). General substance use has been associated with lower
academic achievement in as early as 6
th
grade US students (Sobeck, Abbey et al. 2000;
Ellickson, Tucker et al. 2001). These relationships have been observed among European
adolescents as well with academic underachievers being more likely to have increased
alcohol consumption rates over those with above average academic performance (Miller and
Martin 1999). Indeed a longitudinal study of US adolescents that investigated the
relationships between perceived academic performance and substance use found persistent
perceptions of academic failure at age 13 to predicted alcohol use at the two year follow up
point (Bergen, Martina et al. 2005). Similarly, a 3 year follow up study of 12 year old
children found perceived academic failure to predict heavy alcohol use at age 15 among girls
(Kumpulainen 2002).
Further investigations are required which investigate the reasons that Chinese
adolescent academic achievement seems to be associated with positive changes in alcohol
use behaviors at follow up. We know that academic achievement is a very important aspect
of Chinese adolescent’s lives. Investigations which can determine associations of

123

psychosocial experiences related to academic achievement in Chinese youth will help
determine how these factors may be inter-relating to facilitate alcohol use transition
behaviors.
As indicated no significant interactions were found between any intrapersonal factor
and 30 day smoking, junk food use, and academic score in their relationship to the two
alcohol use levels (monthly use, and binge drinking). No significant interactions were found
between any intrapersonal factor and three parenting variables measured at baseline: parent
monitoring, latchkey status, and parenting style. This may be due to the young age of the
sample and associated low base rates in alcohol use. In addition our measures of
intrapersonal and environmental variable domains may lack the precision to capture these
factors among Chinese youth.
Relatively few studies have examined longitudinal development of adolescent
alcohol use, especially among early adolescents in the west, let alone among a Chinese youth
population. This investigation is among the first to analyze the influence of selected alcohol
use determinants on changes in level of youth of alcohol use at follow up. The findings
presented in this study present definite contributions to our understanding of the factors that
are associated with transition AU behaviors in adolescents. These results represent a unique
contribution to both the fields of youth alcohol use and youth alcohol use among non-
Western adolescents. As reported here we know that adolescents in China start using
alcohol earlier than in most other countries (Hao, Zhonghua et al. 2004). It is critical to
continue to investigating the factors and conditions that lead to alcohol experimentation and
increased consumption behavior change in early ages in this population. Effective youth
alcohol use prevention efforts must consider the potential contributors to this behavior from

124

the most salient domains of an adolescent’s life. The knowledge of the role that potential co-
morbid psychosocial disorders may have on alcohol use behaviors among this population
can help direct investigators to focus intervention efforts on the underlying influences of
alcohol use and on concurrent psychosocial conditions. The marked differences between
results commonly found in Western adolescent populations and those presented with this
sample of Chinese adolescents speaks to the need for further analyses which explore how
distinct socio-cultural and biological factors can influence adolescent alcohol use behaviors.

125

CHAPTER SEVEN: CONCLUSION
The specific aims of this dissertation investigation were to identify intrapersonal and
environmental predictors of alcohol use experimentation, and of transition behaviors to
heavier use in a sample of Chinese adolescents. This study also proposed to investigate the
potential moderating effects that existed between selected intrapersonal and environmental
variables in their relationship to varying levels of alcohol use. Thirdly and most importantly,
this investigation proposed to identify which intrapersonal and environmental domain
variables influenced change in alcohol use consumption behaviors at follow up.
In analyzing the influence of intrapersonal and environmental domain factors on
adolescent alcohol use this study found that it is of critical importance for a complete model
of adolescent alcohol use to study both domains in the same theoretical model. Previous
studies that have tested associations between intrapersonal and environmental factors with
adolescent alcohol use have reported different findings to those presented here where both
domains were tested in the same statistical model. Indeed even studies that I have conducted
in which each domain was tested exclusively showed different associations between each
domain with alcohol use to those presented here. Literature findings and results reported
previously by this investigator have indicated strong associations between each selected
intrapersonal and environmental variable with varying levels of adolescent alcohol use. This
study found that many of these significant associations disappear when analyzing both
domains in the same model. Further investigations are required which analyze possible
effects between intrapersonal and environmental variables in their association with alcohol
use behaviors to see if more refined analyses are required when domain factors are present in
the same model. Investigators must be able to determine if it is necessary to analyze

126

multiple domains when analyzing youth AU determinants. This investigation has
shown that indeed results may vary between studies that analyze both intrapersonal and
environmental domains in a single model and those that investigate these domains
exclusively.
In an attempt to further understand the complex relationships that existed between
intrapersonal and environmental determinants of youth alcohol use this study investigated
the moderating effects of intrapersonal factors on the relationship between environmental
factors and youth alcohol use. This study reported the presence of a number of significant
interactions between intrapersonal and environmental domain factors. The findings
presented in this study supported my general theoretical model that posited the moderating
effects of intrapersonal domain factors on the relationship between environmental factors
and youth alcohol use. The influences of intrapersonal and environmental factors have rarely
been investigated in a sample of Chinese adolescents. The results presented here show that
while many alcohol use determinants are similar to those found in the West, some may be
particular to Chinese youth populations. The moderating effects presented in this study
should serve as a first step in uncovering the complex relationships that exist between the
various determinants that influence Chinese adolescent AU behaviors. The information
presented here supports my view that alcohol use determinants must be analyzed from as
many adolescent life domains as possible given the fact that adolescent alcohol use
behaviors are affected by various life domains concurrently.
Few studies have examined longitudinal development of adolescent alcohol use
behaviors among early adolescents in the West, and fewer still among a Chinese youth
population. This investigation is among the first to analyze the influence of intrapersonal and

127

environmental alcohol use determinants on changes in level of youth of alcohol use at follow
up. This study has contributed to our understanding of the domain factors that are associated
with alcohol use transition behaviors among Chinese adolescents.
Results from the latent growth analysis study presented novel findings that seem
particular to the adolescent Chinese experience. The knowledge of the role that potential co-
morbid psychosocial disorders may have on alcohol use behaviors among this population
can help direct prevention investigators to focus intervention efforts on the underlying
influences of alcohol use and on concurrent psychosocial conditions experienced in this
population. The results presented in study 3 indicate the great need for further analyses
which explore how distinct socio-cultural factors can influence adolescent alcohol use
behaviors.
The three studies included in this dissertation investigation have identified important
determinants of Chinese youth drinking onset and transition behaviors. The first study
provided information on associations of important intrapersonal and environmental domain
factors with varying levels of alcohol use. Important contributions that emerged from this
first study included the fact that junk food use, smoking behavior, and allowance were
incrementally associated with increasing levels of alcohol use in at least one of the analysis
models in both genders. Other important contributions of this study included the findings
that junk food use, smoking behavior and parental drinking were associated with odds of
monthly alcohol use among lifetime users in both genders. Also that depression and smoking
behavior were associated with odds of binge drinking among monthly alcohol users in both
genders.

128

The second study provided information on the moderating effects that intrapersonal
dispositional variables have on the relationship between environmental factors and alcohol
use behaviors. Important contributions to the field of youth alcohol use research,
particularly for the Chinese population, include the findings that indicated that the
associations between the selected intrapersonal variables and both levels of alcohol use
differed according to varying levels of academic score, allowance, latchkey status, and
parental education. Each of these factors relates important cultural practices that seem to be
influencing alcohol use consumption patterns in Chinese youth. More importantly, findings
form this second study demonstrated how these cultural practices may be interacting with
existing dispositional attributes. The findings presented in this second study must function
as a first step to further investigations which identify the role that specific cultural practices
play in influencing both dispositional conditions and youth alcohol use behaviors. Specific
information garnered in such investigations can serve to inform culturally tailored prevention
efforts that take into account detailed aspects of the adolescent’s dispositional “health” and
life environment.
The third study related important findings from latent growth curve analyses which
were utilized to analyze if selected alcohol use determinants influenced changes in level of
youth of alcohol use over a 2 to 3 wave follow up. The findings presented in study three
represent novel contributions to the field of youth alcohol use research. In females results
indicated that monthly alcohol use reduced the experience of high stress at follow up when
30 day smoking was analyzed in the model. If females are in fact utilizing monthly alcohol
use as a stress reduction technique then this information must be assessed and utilized for
prevention efforts. As stated the presence of alternative stress reduction techniques for youth

129

may help delay alcohol use onset in those looking for easy ways to reduce stress. In
contrast, results with males indicated that initial high levels of stress were found to decrease
binge drinking when junk food use and academic score was analyzed in the same model. It
is my conjecture that Chinese adolescent males may be finding efficient ways to cope with
high stress and that these coping mechanisms may be delaying onset of later alcohol use
behaviors. Further investigations are needed which can help clarify the link between high
levels of stress and decreases in alcohol use at follow up. Similarly in males, results showed
that high levels of initial depression influenced a decrease in binge drinking when 30 day
smoking, junk food use, and academic score was present in the analysis respectively.
Further investigations are required which can determine the reasons behind the fact that
initial high levels of depression seem to be influencing decreases in binge drinking in males
at follow up. Detailed investigations which explore both cultural and biological
determinants of adolescent Chinese youth are necessary to explain why initial elevated
dispositional characteristics influence decreases in alcohol use at follow up.
Another contribution to the field of youth alcohol use from this third study are the
findings that showed that initial high academic scores were found to influence change in
monthly drinking in females when stress was present in the model. Also those which showed
that initial high academic scores were found to influence change in binge drinking among
males when hostility was in the model. These findings indicated that initial high academic
scores relate positively to changes in monthly alcohol use and binge drinking behaviors at
follow up. In both cases we also saw that initial high levels of psychosocial factors
positively influenced academic scores at follow up in both genders. These findings are
unique in that in the investigations conducted in the United States and in Europe have

130

shown that poor academic achievement is associated with increased substance use
(Kumpulainen 2002). Further investigations are needed which investigate the reasons that
Chinese adolescent academic achievement seems to be associated with positive changes in
alcohol use behaviors at follow up. Academic achievement investigations that focus on
cultural and psychosocial determinants in Chinese youth can help identify what possible role
these factors play on risky health behaviors such as alcohol use.
Limitations
The cross-sectional nature of the first two studies did not allow me to determine if
any intrapersonal or environmental domain factors were existent before the development of
alcohol use behaviors. The fact that the data for these first two studies was cross-sectional
disallowed any assessment of directionality. We have to consider the possibility that it is
alcohol use behaviors which may be influencing the associations with the selected
intrapersonal and environmental domain variables analyzed in this study. For example, we
do not know with certainty if it is depression that is causing binge drinking behaviors or if
binge drinking behaviors are in fact causing depression. Further studies which look at
longitudinal data in this population may be able to better answer these types of questions.
One other consideration we must make is that the validity of data could be affected
by the self-report nature of the study. There is always the possibility that answers given to
these questionnaires items are not true reflections of the adolescent’s experience because
they do not wish to truthfully disclose substance use and other risky health behaviors. Thus
we must allow for the fact that self reported risky health behaviors may be underreported.
One possible limitation of this study is the measures utilized to assess psychological
variables. Although taken from larger validated measures, the number and breadth of items

131

was abbreviated. However, because we used validated self-report instruments widely used in
the literature I feel confident that I have obtained valid data from the study participants.
Also, given the young age of the sample we may be obtaining associated low base rates in
alcohol use. In addition although all measures went through a thorough process of
translation and back-translation our measures of intrapersonal and environmental variable
domains may lack the precision to capture these factors among Chinese youth.
Finally the generalizability of results to populations other than China must be evaluated
carefully.
Implications
This study was undertaken in order to further explore the field of youth alcohol use to help
investigators in the field gain an increasingly refined understanding of the etiological factors
that contribute to alcohol use onset and how these factors interrelate with one another to
contribute to progression along the gamut of adolescent alcohol use involvement. I believe
that the findings reported in this dissertation contributed to a more detailed understanding of
the factors that influence alcohol use initiation and consumption behavior change. Given
that these findings were attained in a large Chinese youth population lends to the
contributions that this investigation has proposed to the field of youth alcohol use research.
The fact that this study focused on early adolescent Chinese youth, an age and population
that is neglected in alcohol use research is an important contribution to the field. This
investigation had a number of strengths that set it apart from investigations that have
analyzed youth alcohol use in Asian populations. These include the use of youth and parent
report data, a 2 wave follow up, and use of a cohort-sequential latent growth curve analyses
to determine growth in alcohol use rates among youth from grades 7, 8, 10, and 11.

132

Further investigations are required which continue to analyze and differentiate risk
and protective factors for both the onset of alcohol use and those for transition use behaviors.
It is these types of investigations which can best inform prevention efforts in the field by
identifying specific risk factors or combinations of risk and protective factors which may
promote delays of alcohol use onset or curb progression to heavier use along different points
of transition. Additional investigations are required which can further analyze and refine the
information presented in this dissertation investigation to possibly include these findings in
prevention intervention efforts. This study has presented a number of findings which can be
further explored for their value in developing specific prevention efforts which take into
account the daily concurrent influences of an adolescent’s life.

133

BIBLIOGRAPHY
Abrams, D. B. and R. S. Niaura (1987). Social learning theory. Psychological theories of
drinking and alcoholism. H. T. Blane and K. E. Leonard. New York, Guilford: 131–
178.

Anderson, A. R. and C. S. Henry (1994). "Family system characteristics and parental
behaviors as predictors of adolescent substance use." Adolescence 29: 405-420.

Andersson, T., D. Magnusson, et al. (1997). "Early aggressiveness and hyperactivity as
indicators of adult alcohol problems and criminality: a prospective longitudinal
study of male subjects." Studies on Crime and Crime Prevention 6: 7-20.

Atkin, C. K. (1990). "Effects of televised alcohol messages on teenage drinking patterns."
Journal of Adolescent Health Care 11: 10–24.

Barnes, G. M. (1984). "Adolescent alcohol abuse and other problem behaviors: Their
relationships and common parental influences." Journal of Youth and Adolescence
13: 329-348.

Barnes, G. M., A. S. Reifman, et al. (1994). "Longitudinal effects of parenting on alcohol
misuse among adolescents." Alcoholism: Clinical and Experimental Research 18:
507.

Beck, K. H., T. Shattuck, et al. (1999). "Associations between parent awareness, monitoring,
enforcement and adolescent involvement with alcohol." Health Education Research
14: 765-775.

Bergen, H. A., G. Martina, et al. (2005). "Perceived academic performance and alcohol,
tobacco and marijuana use: Longitudinal relationships in young community
adolescents." Addictive Behaviors () 30: 1563–1573.

Boutelle, K. N., D. M. Murray, et al. (2000). "Associations between Exercise and Health
Behaviors in a Community Sample of Working Adults." Preventive Medicine 30:
217–224.

Boyle, M. H., D. R. Offord, et al. (1992). "Predicting substance use in late adolescence:
results from the Ontario Child Health Study Follow-Up." American Journal of
Psychiatry 149: 761-767.

Brook, J. S., C. Nomura, et al. (1989). "A network of influences on adolescent drug
involvement: Neighbourhood, school, peer, and family." Genet. Soc. Gen. Psychol.
Monogr 115: 133-1142.

Brook, J. S., Richter, L. and Rubenstone, E. (2000). "Consequences of adolescent drug use
on psychiatric disorders in early adulthood." Ann. Med 32: 401-407.

134

Burke, J. D., K. C. Burke, et al. (1994). "Increased rates of drug abuse and dependence after
onset of mood or anxiety disorders in adolescence." Hospital and Community
Psychiatry 45: 451-455.

Burt, C. E., L. H. Cohen, et al. (1988). "Perceived family environment as a moderator of
young adolescents' life stress adjustment." American Journal of Community
Psychology 16(1): 101-122.

Buss, A. H. and A. Durkee (1957). "An inventory for assessing different kinds of hostility."
Journal of Consulting Psychology 21: 343-349.

Chao, R. K. (1994). "Beyond parental control and authoritarian parenting style:
Understanding Chinese parenting through the cultural notion of training." Child
Development 65: 1111-1119.

Chao, R. K. (1995). "Chinese and European American cultural models of the self reflected in
mother‘s child-rearing beliefs." Ethos 23: 328–354.

Chao, R. K. and S. Sue, Eds. (1996). Chinese parental influence and their children's success:
A paradox in the literature on parenting styles. Youth and child development in
Chinese societies. Hong Kong, Chinese University of Hong Kong.

Chen, C.-Y. and C. L. Storr (2006). "Alcohol Use and Health-Related Quality of Life among
Youth in Taiwan." Journal of Adolescent Health 39: 752.e9 –752.e16.

Chen, K. and D. B. Kandel (1998). "Predictors and Cessation of Marijuana Use: An Event
History Analysis." Drug and Alcohol Dependence 50: 109-121.

Chilcoat, H. D., T. J. Dishion, et al. (1995). "Parent monitoring and the incidence of drug
sampling in urban elementary school children." American Journal of Epidemiology.
141: 25-31.

Clark, D. B., N. Pollock, et al. (1997). "Gender and comorbid psychopathology in
adolescents with alcohol dependence." Journal of the American Academy of Child
and Adolescent Psychiatry 36: 1195–1202.

Clark, D. B. and K. C. Winters (2002). "Measuring risks and outcomes in substance use
disorders prevention research." Journal of Consulting and Clinical Psychology 70:
1207-1223.
Cochrane, J., H. Chen, et al. (2003). "Alcohol use in China." Alcohol & Alcoholism 38(6):
537–542.

Cohen, D. A. and J. Rice (1997). "Parenting Styles, adolescents substance use and academic
achievement." Journal of Drug Education 27(2): 199-211.

Cohen, S., T. Kamarck, et al. (1983). "A global measure of perceived stress." Journal of
Health and Social Behavior 24: 385-396.

135

Coombs, R. and J. Landsverk (1988). "Parenting Styles and Substance Use during Childhood
and Adolescence." Journal of Marriage and Family 50(2): 473-482.

Cooper, M. L., M. Russell, et al. (1992). "Stress and alcohol use: Moderating effects of
gender, coping and alcohol expectancies." Journal of Abnormal Psychology 101:
139–152.

Crum, R. M., M. E. Ensminger, et al. (1998). "The association of educational achievement
and school dropout with risk of alcoholism: a twenty-five-year prospective study of
inner-city children." Journal of Studies on Alcohol 59(3): 318-9.

Degenhardt, L., W. Hall, et al. (2000). "Cohort trends in the age of initiation of drug use in
Australia." Australian and New Zealand Journal of Public Health 24: 421-426.

Deykin, E. Y., J. C. Levy, et al. (1987). "Adolescent depression, alcohol and drug abuse."
American Journal of Public Health 77: 178–182.

DiClemente, R. J., G. M. Wingood, et al. (2001). "Parental monitoring: Association with
adolescents’ risk behaviors." Pediatrics 107(6): 1363-1368.

Dishion, T. J., G. R. Patterson, et al. (1991). "Family, school, and behavioral antecedents to
early adolescent involvement with antisocial peers." Developmental Psychology 27:
172-180.

Donovan, J. E. and . : (2004). "Adolescent alcohol initiation: A review of psychosocial risk
factors." Journal of Adolescent Health 35: 529.e7–18.

Donovan, J. E. and R. Jessor (1985). "Structure of problem behavior in adolescence and
young adulthood." Journal of Consulting and Clinical Psychology 53: 890-904.

Donovan, J. E., R. Jessor, et al. (1983). "Problem drinking in adolescence and young
adulthood." Journal of Studies on Alcohol 44: 109-137.

Dufour, M. C. (1994). "Are there net health benefits from moderate alcohol consumption?
Morbidity and other parameters of health." Contemp. Drug Probl 21: 163-183.

Duncan, T. E., S. C. Duncan, et al. (1999). An introduction to latent variable growth curve
modeling: Concepts, issues and applications. Mahwah, NJ, Erlbaum.

Eccles, J. S. and B. L. Barber (1999). "Student council, volunteering, basketball, or marching
band: What kind of extracurricular involvement matters?" Journal of Adolescent
Research 14: 10-43.

Ellickson, P. L., J. S. Tucker, et al. (2001). "High-risk behaviors associated with early
smoking: Results from a 5-year follow-up." Journal of Adolescent Health 28(6):
465–473.

136

Elliott, D. S., D. Huizinga, et al. (1985). Explaining delinquency and drug use. Beverly Hills,
CA, Sage.

El-Sayed, M., N. Ali, et al. ( 2005). "Interaction Between Alcohol and Exercise
Physiological and Haematological Implications." Sports Med 35(3): 257-269.

Everett, S., Snuth, R. and Tribble, J. (1998). "Tobacco and alcohol use in top-grossing
American films." Journal of Community Health 23: 317–324.

Formoso, D., N. A. Gonzales, et al. (2000). "Family Conflict and Children's Internalizing
and Externalizing Behavior: Protective Factors." American Journal of Community
Psychology 28(2): 175-199.

Freisthler, B., P. J. Gruenewald, et al. (2003). "Evaluating alcohol access and the alcohol
environment in neighborhood areas." Alcoholism - Clinical and Experimental
Research 27(3): 477-484.

Fromme, K., E. Stroot, et al. (1993). "Comprehensive effects of alcohol: Development and
psychometric assessment of a new expectancy questionnaire." Psychological
Assessment 5: 19-26.

Furnham, A., Ingle, H., Gunter, B. (1997). "A content analysis of alcohol portrayal and
drinking in British television soap operas." Health Education Research 12: 519–529.

Getz, J. G. and J. H. Bray (2005). "Predicting Heavy Alcohol Use Among Adolescents."
American Journal of Orthopsychiatry 75(1): 102-116.

Gilman, S. E. and H. D. Abraham (2001). "A longitudinal study of the order of onset of
alcohol dependence and major depression." Drug & Alcohol Dependence 63: 277-
286.

Glantz, S. A. and L. L. Mandel (2005). "Since school-based tobacco prevention programs do
not work, what should we do?" Journal of Adolescent Health 36(3): 157-159.
Grant, B. F. and D. A. Dawson (1997). "Age At Onset Of Alcohol Use And Its Association
With DSM-IV Alcohol Abuse and Dependence: Results from the National
Longitudinal Alcohol Epidemiologic Survey." Journal of Substance Abuse 9: 103-
110.

Grant, B. F. and T. C. Harford (1995). "Comorbidity between DSM-IV alcohol use disorders
depression: results of a national survey and major depression: results of a national
survey." Drug and Alcohol Dependence 39: 197-206.

Greeley, J. and T. Oei (1999). Alcohol and tension reduction. Psychological theories of
drinking and alcoholism. K. E. Leonard and H. T. Blane. New York, Guilford: 14-
53.

137

Griffin, K. W., G. J. Botvin, et al. (2003). "Common predictors of cigarette smoking, alcohol
use, aggression, and delinquency among inner-city minority youth." Addictive
Behaviors 28: 1141–1148.

Griffin, K. W., L. M. Scheier, et al. (2000). "Ethnic and gender differences in psychosocial
risk, protection, and adolescent alcohol use." Prevention Science 1: 199-212.

Guinn B, S. T., Jorgenson L. (1997). "Mexican American female adolescent self-esteem: the
effect of body image, exercise behavior, and body fatness." Hisp J Behav Sci 19(4):
517-526.

Hao, W., H. H. Chen, et al. (2005). "China: alcohol today." Addiction 100: 737–741.

Hao, W., D. S. Young, et al. (1995). "Alcohol drinking in China: present, future and policy."
Chinese Journal of Clinical Psychology 3: 243-248.

Hao, W., D. S. Young, et al. (1998a). "Alcohol drinking and drinking related problems in six
areas in China: Part I. Methodology and alcohol drinking in general population."
Chinese Journal of Clinical Psychology 6: 65-70.

Hao, W., S. Zhonghua, et al. (2004). "Drinking and Drinking Patterns and Health Status in
the General Population of Five Areas of China." Alcohol & Alcoholism 39(1): 43-
52.

Hartos, J. L. and T. G. Power (2000). "Relations Among Single Mothers' Awareness of Their
Adolescents' Stressors, Maternal Monitoring, Mother-Adolescent Communication,
and Adolescent Adjustment." Journal of Adolescent Research 15(5): 546-563.

Hawkins, J. D., J. F. Catalano, et al. (1992). "Risk and Protective Factors for Alcohol and
other Drug Problems in Adolescence and Early Adulthood: Implications for
Substance Abuse Prevention." Psychological Bulletin 112: 64-105.
Hawkins, J. D. and J. G. Weis (1985). "The social development model: An integrated
approach to delinquency prevention." Journal of Primary Prevention 6: 73-97.

Hebbell, B., B. Andersson, et al. (2003). ESPAD Report 2003: Alcohol and Other Drug Use
Among Students in 35 European Countries. Copenhagen, The European School
Survey Project on Alcohol and Other Drugs (ESPAD).

Henry, C. S., L. C. Robinson, et al. (2003). "Adolescent perceptions of their family system,
parents' behavior, self-esteem, and family life satisfaction in relation to their
substance use." Journal of Child & Adolescent Substance Abuse 13(2): 29-59.

Herzog, D. B., Dorer, D.J., Keel, P.K., Selwyn, S.E., Ekeblad, E.R., & Flores, A.T. (1999).
"Recovery and relapse in anorexia and bulimia nervosa: A 7.5-year follow-up
study." Journal of the American Academy of Child and Adolescent Psychiatry 38:
829–837.

138

Hughes, J. R. (2006). "Should criteria for drug dependence differ across drugs?" Addiction
101(s1): 134–141.

Hussong, A. M. (2003). "Further refining the stress-coping model of alcohol involvement."
Addictive Behaviors 28: 1515–1522.

Institute of Alcohol Studies (2003). Young People, Alcohol & Other Drugs-International
Comparisons. The Quay St Ives, Institute of Alcohol Studies.

Jackson, C., L. Henriksen, et al. (1997). "The early use of alcohol and tobacco: Its relation to
children's competence and parents' behavior." American Journal of Public Health
87(3): 359-364.

Jackson, K. M., K. J. Sher, et al. (2005). "Conjoint Developmental Trajectories of Young
Adult Alcohol and Tobacco Use." Journal of Abnormal Psychology 114(4): 612–
626.

Janghorbani, M., S. Y. Ho, et al. (2003). "Prevalence and correlates of alcohol use: a
populationbased study in Hong Kong." Addiction 98: 215–224.

Jansen, R. E., H. E. Fitzgerald, et al. (1995). "Pathways into risk: Temperament and behavior
problems in three- to five-year-old sons of alcoholics." Alcohol Clin Exp Res 19:
501–509.

Jernigan, D. H. (2001). Global Status Report: Alcohol and Young People. Geneva, World
Health Organization: 1-44.

Jessor, R. and S. L. Jessor (1977). Problem behavior and psychosocial development. New
York, Academic Press.

Johnson, V. and R. J. Pandina (1991). "Effects of the family environment on adolescent
substance use, delinquency, and coping style." American Journal of Drug Alcohol
Abuse 17: 71-88.

Johnston, L. D., P. M. O'Malley, et al. (2005). Monitoring the Future national results on
adolescent drug use: Overview of key findings, 200. Bethesda, MD, National
Institute on Drug Abuse.

Kandel, D. B., V. H. Raveis, et al. (1991). "Suicidal ideation in adolescence: depression,
substance use, and other risk factors." Journal of Youth and Adolescence 20: 289-
309.

Karatzias, A., K. G. Power, et al. (2001). "Predicting Use and Maintancence of Use of
Substances in Scottish Adolescents." Journal of Youth and Adolescence 30(4).

139

Kessler, R. C., R. M. Crum, et al. (1997). "Lifetime co-occurrence of DSM-IIIR alcohol
abuse and dependence with other psychiatric disorders in the National Comorbidity
Survey." Archives of General Psychiatry 54: 313–321.

Kessler, R. C., C. B. Nelson, et al. (1996). "Comorbidity of DSM-III-R major depressive
disorder in the general population: results from the US National Comorbidity
Survey." British Journal of Psychiatry-Supplement: 17–30.

Kumpfer, K. L. and C. W. Turner (1990-1991). "The social ecology model of adolescent
substance abuse: Implications for prevention." International Journal of the
Addictions(25): 435-463.

Kumpulainen, K. (2000). "Psychiatric symptoms and deviance in early adolescence predict
heavy alcohol use three years later." Addiction 95: 1847–1857.

Kumpulainen, K. (2002). "Depressive symptoms at the age of 12 years and future heavy
alcohol use." Addictive Behaviors 27: 425–436.

LeDoux, S., P. Miller, et al. (2002). "Family structure, parent-child relationships, and
alcohol and other drug use among teenagers in France and the United Kingdom."
Alcohol and Alcoholism 37(1): 52-60.

Lenzenweger, M. F., M. D. Johnson, et al. (2004). "Individual Growth Curve Analysis
Illuminates Stability and Change in Personality Disorder Features The Longitudinal
Study of Personality Disorders." Arch Gen Psychiatry 61: 1015-1024.

Levy, D. T., T. R. Miller, et al. (1999). Costs of Underage Drinking. Calverton, MD, Pacific
Institute For Research and Evaluation.

Lewinsohn, P. M., P. Rohde, et al. (1999). "Level of current and past adolescent smoking as
predictors of future substance use disorders in young adulthood." Addiction 94:
913–921.

Li, X., X. Fang, et al. (2003). "Parental monitoring among adolescents in Beijing, China."
Journal of Adolescent Health 33(2): 130-132.

Lin, Y. P. and T. J. Cheng (2002). "Why can’t Chinese Han drink alcohol? Hepatitis B virus
infection and the evolution of acetaldehyde dehydrogenase deficiency." Medical
Hypotheses 59(2): 204–207.

Loukas, A. and H. M. Prelow (2004). "Externalizing and Internalizing Problems in Low-
Income Latino Early Adolescents: Risk, Resource, and Protective Factors." Journal
of Early Adolescence 24(3): 250-273.

Maggs, J. L., J. Schulenberg, et al., Eds. (1997). Developmental transitions during
adolescence: Health promotion implications. Health risks and developmental
transitions during adolescence. New York, Cambridge University Press.

140

Mayhew, K. P., B. R. Flay, et al. (2000). "Stages in the Development of Adolescent
Smoking." Drug and Alcohol Dependence 59: S61-S82.

Melnick, M. J., K. E. Miller, et al. (2001). "Tobacco use among high school athletes and
non-athletes: Results of the 1997 Youth Risk Behavior Survey." Adolescence 36:
727-747.

Miller, P. and P. Martin (1999). "Truancy and Percieved School Performance: An Alcohol
and Drug Study of UK Teenagers." Alcohol & Alcoholism 34(6): 886-893.

Moss, H. B. and K. G. Lynch (2001). "Comorbid disruptive behavior disorder symptoms and
their relationship to adolescent alcohol use disorder." Drug and Alcohol Dependence
64: 75–83.

Mulhall, P. F., D. Stone, et al. (1996). "Home Alone: Is It a Risk Factor for Middle School
Youth and Drug Use?" Journal of Drug Education 26(1): 39-48.

National Health and Medical Research Council, N. (2001). Australian Alcohol Guidelines:
Health Risks and Benefits. Canberra, National Health and Medical Research
Council, (NHMRC).

Neumark-Sztainer D, S. M., Hannan PJ, Perry CL, Irving LM. (2002). "Weight-related
concerns and behaviors among overweight and non-overweight adolescents." Arch
Pediatr Adolesc Med 156: 171 -178.

Newcomb, M. D., P. M. Bemtler, et al. (1986). "Alcohol use and dissatisfaction with self and
life: A longitudinal analysis of young adults." Journal of Drug Issues 16(4): 479-
494.

Nieri, T., S. Kulis, et al. (2005). "Body image, acculturation, and substance abuse among
boys and girls in the Southwest." American Journal of Drug and Alcohol Abuse
31(4): 617(23).

Petraitis, J., B. R. Flay, et al. (1995). "Reviewing Theories of Adolescent Substance Use:
Organizing Pieces in the Puzzle." Psychological Bulletin 117(1): 67-86.

Power, T. G., D. Cynthia, et al. (2005). "Predicting patterns of adolescent alcohol use: a
longitudinal study." Journal of Studies on Alcohol 74(8): 66-71.

Radloff, L. S. (1977). "The CES-D Scale: A self-report depression scale for research in the
general population." Journal of Applied Psychological Measures 1(3): 385-401.

Read, J. and R. A. Brown (2003). "The Role of Physical Exercise in Alcoholism Treatment
and Recovery." Professional Psychology: Research and Practice 34(1): 49–56.

141

Regier, M. E., D. S. Farmer, et al. (1990). "Comorbidity of mental disorders with alcohol
and other drug abuse." JAMA 264: 2511–2518.

Riala, K., I. Isohaaai, et al. (2003). "Low Eduacational Performance is Associated With
Drunk Driving: A 31 year Follow-Up of the Northern Finland 1966 Birth Cohort."
Alcohol & Alcoholism 38(3): 219–223.

Richardson, J. L., K. Dwyer, et al. (1989). "Substance Use Among Eighth-Grade Students
Who Take Care of Themselves After School." Pediatrics 84: 556-566.

Rosal, M. C., J. K. Ockene, et al. (2000). "Prevalence and Co-occurrence of Health Risk
Behaviors among High-Risk Drinkers in a Primary Care Population." Preventive
Medicine 31: 140–147.

Rose, R. J. (1998). "A Developmental Behavior-Genetic Perspective on Alcoholism Risk."
Alcohol Health & Research World 22(2): 131-143.

Sabrina Oesterle, K. G. H., J. David Hawkins, Jie Guo, Richard F. Catalano and Robert D.
Abbott. (2004). "Adolescent heavy episodic drinking trajectories and health in
young adulthood." Journal of Studies on Alcohol 65(2): 204-9.

Sen, B. (2002). "Does alcohol use increase the risk of sexual intercourse among
adolescents? Evidence from the NLSY97." Journal of Health Economics. 21: 1085–93.

Sher, K. J. (2000). "High comorbidity of alcohol, mental disorders warrants new approaches:
Current classification systems may hinder science." APS Observer 13(9): 16-17.

Sher, K. J., E. Grekin, et al. (2005). "The development of alcohol problems." Annual Review
of Clinical Psychology 1: 493-523.

Simons, R. L., R. D. Conger, et al. (1988). "A multistage social learning model of the
influences of family and peers upon adolescent substance abuse." Journal of Drug
Issues 18: 293-315.

Simons-Morton, B., R. Chen, et al. (2004). "Latent Growth Curve Analyses of Peer and
Parent Influences on Smoking Progression Among Early Adolescents." Health
Psychology 23(6): 612–621.

SLATER, M. D., M. D. BASIL, et al. (1999). "A Cluster Analysis of Alcohol-Related
Attitudes and Behaviors in the General Population." Journal of Studies on Alcohol
60(5): 667.

Sobeck, J., A. Abbey, et al. (2000). "Predicting early adolescent substance use: Do risk
factors differ depending on age of onset?" Journal of Substance Abuse 11(1): 89–
102.

142

Steinberg, L., A. Fletcher, et al. (1994). "Parental monitoring and peer influences on
adolescent substance use." Pediatrics 93: 1-5.

Swaim, R. C., J. L. Deffenbacher, et al. (2004). "Concurrent and Prospective Effects of
Multi-Dimensional Aggression and Anger on Adolescent Alcohol Use." Aggresive
Behavior 30: 356–372.

THOMSEN, S. R. and D. REKVE (2006). "The relationship between viewing US-produced
television programs and intentions to drink alcohol among a group of Norwegian
adolescents." Scandinavian Journal of Psychology 47: 33–41.

Thoren, P., J. S. Floras, et al. (1990). "Endorphins and exercise: Physiological mechanisms
and clinical implications." Medicine and Science in Sports and Exercise 22: 417–
428.

Tosteson, T. D., J. S. Buzas, et al. (2003). "Power and sample size calculations for
generalized regression models with covariate measurement error." Statistics in
Medicine 22(7).

Unger, J. B., Y. L. Chen, et al. (2001). "Adolescent smoking in Wuhan, China: Baseline data
from the Wuhan Smoking Prevention Trial." American Journal of Preventive
Medicine 21(3): 162-169.

Valois, R. E., M. L. Vincent, et al. (1993). "Adolescent risk behaviors and the potential for
violence: A look at what's coming to campus." Journal of the Association for
College Health 41: 141-143.

VAN DEN BULCK, J. and K. BEULLENS (2005). "Alcohol & Alcoholism." TELEVISION
AND MUSIC VIDEO EXPOSURE AND ADOLESCENT ALCOHOL USE
WHILE GOING OUT 40(3): 249–253.

W.H.O. (2004). Global Status Report on Alcohol 2004. Geneva, WHO.

Webb, J. A., J. H. Bray, et al. (2002). "Gender, Perceived Parental Monitoring, and
Behavioral Adjustment: Influences on Adolescent Alcohol Use." American Journal
of Orthopsychiatry 72(3): 392-400.

Weiner, M. D., M. A. Pentz, et al. (2001). "From early to late adolescence: Alcohol use and
anger relationships." Journal of Adolescent Health 28: 450–457.

Werch, C., M. J. Moore, et al. (2005). "A multihealth behavior intervention integrating
physical activity and substance use prevention for adolescents." Prevention Science
6(3): 213 - 14.

White, H. R., J. Brick, et al. (1993). "A longitudinal investigation of alcohol use and
aggression in adolescence." Journal of Studies on Alcohol, Supplement Supplement
No. 11: 62-77.

143

Wills, T. A. and E. Hirky, Eds. (1996). Coping and substance abuse: A theoretical model and
review of the evidence. Handbook of coping: Theory, research and applications.
New York, Wiley.

Windle, R. C. and M. Windle (1995). " Longitudinal patterns of physical aggression:
associations with adult social, psychiatric, and personality functioning and
testosterone levels." Developmental Psychology 7: 563-585.

Xing, Y., C. Ji, et al. (2006). "Relationship of Binge Drinking and Other Health-
Compromising Behaviors among Urban Adolescents in China." Journal of
Adolescent Health 39: 495–500.

Yoon, Y. H., H. Yi, et al. (2003). Accidental* Alcohol Poisoning Mortality in the United
States, 1996–1998. Rockville, MD, National Institute on Alcohol Abuse and
Alcoholism.

Zhou, X., Z. Su, et al. (2006). "A comparative survey on alcohol and tobacco use in urban
and rural populations in the Huaihua District of Hunan Province, China." Alcohol
39: 87e96.

Zweig, J. M., S. D. Phillips, et al. (2002). "Predicting Adolescent Profiles of Risk: Looking
Beyond Demographics." Journal of Adolescent Health 31(4): 343-353.

Appendix: Table A. Review of Empirical Studies of Adolescent Alcohol use Determinants*.

Study Population Country Type of Study Substances
1
Chassin, Pillow, Curran, Molina,
Barrera, 1993
454 adolescents 10.5 to 15.5 yrs US Cross-sectional Alcohol
Tobacco
Marijuana
2
Killen, Hayward, Wilson, Haydel, et al,
1996
1,164 9th grade students US Longitudinal 1 wave Alcohol

44.4% Wh; 21.3% A; 17.4% H; 6.0%
PI; 3.4% AfAm; 1.1% NA; 6.4% Oth.

3 Chassin, Curran, Hussong, Colder, 1996 454 adolescents 10.5 to 15.5 yrs US Longitudinal 3 waves Alcohol
Tobacco
78% Wh; 21% H Marijuana
4
Li, Fang, Stanton, Feigelman, Dong,
1996
1,040 6-8-10 grade students China Cross-sectional Alcohol
Tobacco
5 Duncan, Alpert, Duncan, Hops, 1997 480 adolescents mean age: 13 yrs US Longitudinal 5 waves Alcohol
92% Wh
6 Colder, Chassin, 1997 427 adolescents 12-18 years of age US Cross-sectional Alcohol
70% Wh
7 Marcoux, Shope, 1997 3,946 5th - 8th grade students US Longitudinal 1 wave Alcohol
primarily white
8 Dishion, Capaldi, Yoerger, 1999 206 youth 10 yrs US Longitudinal cohort Alcohol
Tobacco
Marijuana

144

Table A. continued

Study Population Country Type of Study Substances
9 Chassin, Pitts, DeLucia, Todd, 1999 246 adolescents 10.5 - 15.5 yrs US Longitudinal 4 waves Alcohol
23.4% H Marijuana/Hashish
Cocaine/Crack
Tranquilizers
Barbiturates
Amphetamines
Hallucinogens
10 Wagner, Myers, McIninch, 1999 332 9-12 grade students US Cross-sectional Alcohol
Marijuana/Hashish
Cocaine/Crack
11
Beck, Shattuck, Haynie, Crump, Simons-
Morton, 1999 454 adolescents 14-19 yrs US Cross-sectional Alcohol
76.5% Wh; 17% AfAm; 6.5% Oth
12 Barnes, Reifman, Farrell, Dintcheff, 2000 699 adolescents 13-16 yrs US Longitudinal 6 waves Alcohol
29% AfAm; 71% Wh, Oth
13 Scheier, Botvin, Griffin, Diaz, 2000 740 7th - 10th grade stuents US Longitudinal (3 wave) Alcohol
90% White Tobacco
Marijuana
14 Griffin, Botvin, Scheier, Diaz, Miller, 2000 228 6th grade students US Cross-sectional Alcohol

88% AfAm; 2% H; 2% As; 1%
Wh; 7% Oth Tobacco

145

Table A. continued

Study Population Country Type of Study Substances
15 Kelder, Orpinas, McReynolds, 2001 5721 adolescents 6, 7, 8 grade students US Cross-sectional Alcohol
62% H Tobacco
Marijuana
16 Brody, Ge, 2001 120 12 yr olds US Longitudinal 3 waves Alcohol
100% Caucasian
17
Makini, Hishinuma, Kim, Carlton,
et al, 2001 2980 high school adolescents Hawaii Cross-sectional Alcohol
62.21% Hawaiian Tobacco
Marijuana
Cocaine
Inhalant use
18 Guo, Hawkins, Hill, Abbott, 2001 808 5th grade students US Longitudinal 3 waves Alcohol
46% Wh; 24% AfAm; 21% As; 6% NA; 3% Oth Tobacco
Marijuana
19
Lonczak, Huang, Catalano,
Hawkins, Hill, et al, 2001 808 5th grade students US Longitudinal 8 waves Alcohol
46% Wh; 24% Af Am; 21% As
20 Moss, Lynch, 2001 805 adolescents mean age: 16 yrs US Cross-sectional Alcohol
77% Wh; 16.14% AfAm; 4.84% Oth
21 Epstein, Williams, Botvin, 2002 3212 7 grade students US Cross-sectional Alcohol
AfAm 71% Caribbean/West Indian 29% Tobacco
Marihuana

146

Table A. continued

Study Population Country Type of Study Substances
22
Griffin, Botvin, Scheier, Doyle,
Williams, 2002 5,536 6th grade students US Cross-sectional Alcohol

41% Af Am; 32% H; 9% Wh; 5% As; 2% AI;
and 11% biracial Tobacco
23
Colder, Campbell, Ruel,
Richardson, Flay, 2002 1,918 7, 8, 9, 12 grades students US Longitudinal 5 waves Alcohol

41.8% H; 32.2% Wh; 9.6% As; 10.4% Af Am;
5.9% Oth
24 Kumpulainen, Roine, 2002 1,316 8 yr old children Finland longitudinal 3 waves Alcohol
Tobacco
25 Chassin, Pitts, Prost, 2002 238 children of alcoholics and 208 controls. US Longitudinal 3 waves Alcohol
mean age: 13 yrs Tobacco
77.2% Wh Marijuana
26
Koposov, Ruchkin, Eisemann,
Sidorov, 2002 387 adolescents mean age 15.6 Russia Cross-sectional Alcohol
27
Wennberg, Andersson, Bohman,
2002 212 birth cohort US
Longitudinal birth
cohort Alcohol
28 Webb, Bray, Getz, Adams, 2002 1,672 7th - 10th grade students US Longitudinal 2 wave Alcohol
Equal proportions of Wh, Af Am, H
29 Adalbjarnardottir, Rafnsson, 2002 1,293 14 year-old students Iceland Longitudinal 1 wave Alcohol
Tobacco
Marijuana
147

Table A. continued

Study Population Country Type of Study Substances
30
Tucker, Orlando, Ellickson,
2003 5,694 individuals age 13 US Longitudinal 6 waves Alcohol use
Marijuana
Cocaine
31 Unger, Sussman, Dent, 2003 966 continuation high school students US Cross-sectional Alcohol
Tobacco
Marijuana
32
Li, Fang, Stanton, Su, Wu,
2003 323 7th- 9th grade students China Cross-sectional Alcohol
Tobacco
33
Griffin, Botvin, Nichols,
Scheier, 2004 2,229 7th - 9th grade students US Longitudinal 2 waves Alcohol
48% AfAm, 32% H, 9% As, 5% Wh
34
Obando, Kliewer, Murrelle,
Svikis, 2004 5,268 adolescents, 12–20 yrs Costa Rica Cross-sectional Alcohol use

35
Goodwin, Fergusson,
Horwood, 2004 1,265 children New Zealand Longitudinal birth cohort Alcohol use
Tobacco
Marijuana
36
Swaim, Deffenbacher,
Wayman, 2004 2683 7th - 12th grade students US Cross-sectional Alcohol Use
57.94% H; 42% Wh Prospective

148

Table A. continued

Study Population Country Type of Study Substances
37 Beck, Boyle, Boekeloo, 2004 406 adolescents 12-17 yrs age US Longitudinal 1 wave Alcohol
79.3% AfAm
38 Catanzaro, Laurent, 2004 210 9–12 grade students US Cross-sectional Alcohol
87.1% Wh; 3.8% Af Am; 2.4% As;
39 Clark, Thatcher, Maisto, 2004 361 adolescents 14 - 17 yrs US Cross-sectional Alcohol
87.5% Wh; 7.8% Af Am; 4.7% Oth Marijuana
Cocaine
40 Nash, McQueen, Bray, 2004 3,620 9-12 grade students US Longitudinal 3 waves Alcohol
27% Wh; 22% Af Am; 40% H; 11%
41 Callas,Flynn, Worden, 2004 2,919 7 - 8 grade students US Cross-sectional Alcohol
Tobacco
42 Lee, Akers, Borg, 2004 3,065 7 - 12 grade students US Cross-sectional Alcohol
98% Wh Marihuana
43 Epstein, Griffin, Botvin, 2004 1,459 8th-12th grade students US Longitudinal 2 waves Alcohol
54% H; 20%AfAm; 16% Wh; 7% A
44
Bergen, Martin, Roeger,
Allison, 2005 2,603 adolescents mean: 13 yrs Australia Longitudinal 3 waves Alcohol
Tobacco
45 Dent, Grube, Biglan, 2005 16,694 students 16–17 yrs US Repeated Cross-sectional Alcohol
4% As; 8% H; 85% Wh
149

Table A. continued

Study Population Country Type of Study Substances
46 Kuntsche, Kuendig, 2005
1,216 9th graders mean age = 15.3
yrs Switzerland Cross-sectional Alcohol
47 Getz, Bray, 2005 3,675 middle school age students US Longitudinal 3 waves Alcohol
39% Wh; 26% Af Am; 35% H Marijuana
48
Power, Stewart, Hughes,
Arbona, 2005 1,253 9 - 12 grades students US Longitudinal 4 waves Alcohol

38.78 % Wh; 33.91 AfAm; 27.29%
H
49 Chen, Storr, 2006 2235 adolescents 12–18 yrs China (Taiwan) Cross-sectional retrospective Alcohol
Tobacco
50 Duncan, Duncan, Strycker, 2006 405 youth 9, 11, 13 yrs US Longitudinal cohort Alcohol
50.4% AfA, 49.6% Wh.
51 Xing, Ji, Zhang, 2006 54,040 7 - 12 grade students China Cross-sectional Alcohol use
Tobacco
52 Yeh, Chiang, Huang, 2006 779 high school students Taiwan Cross-sectional Alcohol
56.7% Han; 43.3% aboriginals
53
Pardini, White, Stouthamer-
Loeber, 2007 506 7 grade students US Cross-sectional Alcohol
56% Af Am; 41% Wh
54
Griffin, Botvin, Epstein, Doyle,
Diaz, 2007 1,132 7th grade students US Longitudinal 3 waves Alcohol
91% Wh

*Note: Studies were included in this table review if (1) the study population focused on children and adolescent populations; (2) empirically
identified and tested at least one of the following: (a) risk factors relating to onset of alcohol use; (b) risk factors to heavy alcohol use
or binge drinking; (c) assessed alcohol-related outcome variables; (d) included objective or self-report measures of drinking behavior.

150

Abstract (if available)

Abstract This study investigated intrapersonal and environmental determinants of alcohol use in a Chinese youth population to explore the role that such determinants have on alcohol use behaviors. This study analyzed which determinants functioned as risk and protective factors for the transition to increased alcohol use

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California

PDF

Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents

PDF

Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults

PDF

Factors and correlates of sexual behaviors among young adults from continuation high schools

PDF

Stage of readiness and learning styles -- a tailored HIV/AIDS prevention intervention for youth detained in the Los Angeles County Juvenile Justice system

PDF

Exposure to alcohol advertising on television and alcohol use among young adolescents

PDF

Motivation and the meanings of health behavior as factors associated with eating behavior in Latino youth

PDF

Sociocultural stress, coping and substance use among Hispanic/Latino adolescents

PDF

Genetic variants and smoking progression in Chinese adolescents

PDF

Non-steroidal anti-inflammatory drugs, anti-acid drugs and passive smoking as impact factors for esophageal and gastric adenocarcinomas

PDF

The role of acculturation in the sun-safe behaviors of US Latino adults in the United States

PDF

Objectively measured physical activity and related factors in minority youth

PDF

Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health

PDF

Associations between sex steroid hormones, hemostatic factors and atherosclerosis

PDF

Exploring the role of peer influence, linguistic acculturation, and social networks in substance use

PDF

Anxiety symptoms and nicotine use among adolescents and young adults

PDF

A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use

PDF

The effect of an educational fotonovela on the prevention of secondhand smoke

PDF

Correlates of compliance with mammography screening guidelines among low-income Latinas: An exploratory study

PDF

Competing risks: the role of the perceived consequences of refusing to share injection equipment among injection drug users

Asset Metadata

Creator Ortega, Enrique (author)

Core Title Intrapersonal and environmental factors associated with Chinese youth alcohol use experimentation and binge drinking behaviors

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Publication Date 07/25/2008

Defense Date 03/28/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Adolescents,alcohol use,china,determinants,OAI-PMH Harvest

Place Name China (countries)

Language English

Advisor Johnson, Carl Anderson (committee chair), Baezconde-Garbanati, Lourdes (committee member), Chi, Iris (committee member), Sun, Ping (committee member), Unger, Jennifer B. (committee member)

Creator Email enriqueo@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1412

Unique identifier UC1226414

Identifier etd-Ortega-20080725 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-198550 (legacy record id),usctheses-m1412 (legacy record id)

Legacy Identifier etd-Ortega-20080725.pdf

Dmrecord 198550

Document Type Dissertation

Rights Ortega, Enrique

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu