Association Between Mental Health problems, Substance Use, and Social Support in Youth

Association Between Mental Health problems,

Substance Use, and Social Support in Youth

Chapter 1: Introduction to the Study

Mental health problems are a major public health concern and may be associated

with substance use (Conway et al., 2018) and inadequate social support over time

(Holden, Dobson, Ware, Hockey, & Lee, 2015). Risk factors for mental health problems

are widely considered to include genetic-environmental and cognitive factors (Cheng et

al., 2014) and may be prevalent not only in the adult population but also among students

before they reach their adult years. Mental health problems among young adults, in

particular, may be associated with the level of social support available to the individual

and may also be associated with substance use (Cheng et al., 2014). Substance use in this

study referred to the use of legal or illegal substances without a prescription (Birkeland,

Weimand, Ruud, Hoie, & Vederhus, 2017). Manwell et al. (2015) define mental health

problems as the absence of mental disease, which includes the biological, psychological,

and social factors that could affect a person's mental state and their function in society.

Social support is a formal or informal relationship that an individual has with another

person (Joni & Leonard, 2013). Social support that comes from family members, friends,

and peers is considered informal social support, whereas when it comes from

organization and healthcare professionals, it is considered formal social support (Joni &

Leornard, 2013). Substance use and lack of social support can impact the young adult’s

sense of well-being and may result in mental health problems, which could create

substantial societal burdens.

Existing literature suggests that mental health is associated with substance use

(Amosu, Onifade, & Adamson, 2016; Conway et al., 2018) and social support (Milner,

Krnjacki, & LaMontagne, 2016) may vary across the overall population and affects all

age groups including the young (Milner et al., 2016). Substance use is a national

epidemic and a public health crisis that has increased in prevalence over time and places

the population affected at risk for early mortality and morbidity (Hopkins, Landen, &

Toe, 2018).

By examining the relationship between mental health problems, substance use,

and social support, for young adults before they enter their adult years, specific and more

focused interventions may be developed to promote mental health problems strategies

that may improve outcomes among the young population. If the correlates of mental

health are identified early in a young person’s life, interventions may be implemented

early, and thus help the youth with mental health problems to become more productive

members of the society (Costello, 2016) and thereby promote positive social change.

Chapter 1 of this dissertation includes the background of the study, problem statement,

purpose of the study, research questions and hypotheses, conceptual framework, nature of

the study, definitions, assumptions, scope and delimitations, limitations, the significance

of the study, significance to social change and the summary.

Background of the Study

Studies have been conducted examining the association between mental health

problems, substance use, and social support in the overall adult population (Amosu et al.,

2016; Firestone et al., 2015; Jibeen, 2016; Ni, Harrington, Wilkins-Turner, 2017; Zhuang

& Wong, 2017). These studies were conducted in schools, communities, healthcare

facilities, and prisons and are the basis for interventions currently used for the population

suffering from mental illness such as identifying youths who are vulnerable to ensure

prevention and put in place early interventions which increased health outcomes. Further

interventions and treatments addressing mental health problems are needed (Conway et

al., 2018) to ensure appropriate treatment and management of care.

Despite many studies on social support and substance use associated with mental

health in the overall population, there is a lack of research on social support, substance

use, and the relationship to mental illness or mental health problems among the young

(Cheng et al., 2014; Jibeen, 2016; Lakey, Vander Molen, Fles, & Andrews, 2016; Lerissa

et al., 2017; Levula, Wilson, & Harre, 2016; Mason, Zaharakis, & Benotsch, 2014).

Shahdadi,Mansouri, Nasiri, & Bandani, (2017) examined the association between

mental health and social support among university students and found an increase in the

number of university students experiencing mental health problems. Another study

conducted by Ni et al. (2017) showed that certain age groups, especially youths, were at a

higher risk for mental health problems following substance use such as drinking and

illicit drug use. Mason et al. (2014) and Shahdadi et al. maintained that there is a

relationship between social support, substance use, and mental health problems and that

this should be examined to identify potential risk factors. The authors found that

substance use among young adults has increased without a corresponding increase in

social support; therefore, it is necessary to conduct research on substance use, social

support, and mental health problems among youth. For this study, I examined the

relationship between mental health problems and substance use and social support by

using secondary data obtained from middle school (MS) and high school (HS) students

in the Maine Integrated Youth Health Survey (MIYHS; 2017).

Substance use is associated with social support. Individuals who have

compromised social support such as family and friends may experience the consequence

of substance use that may affect their well-being (Staton, Royse, & Leukfeld, 2007).

Because social support is considered a buffer for an individual who experiences life crises

(University of Minnesota, 2016), individuals who have adequate social support are less

likely to begin substance use in times of distress. This is because positive social support is

an enhancer of positive behavior that is needed to avert drug use.

Substance use is associated with mental health problems such as anxiety,

depression, and suicide (Amosu et al., 2016; Lerrisa et al., 2017; Mason et al., 2014).

Early use of the substance may lead to many health risks, which include mental health

problems (Henchoz et al., 2016). The use of substances could impact some physiological

pathways that manifest as untoward health consequences. For instance, the use of

nicotine-containing substances could lead to addiction, and addiction is known to be

associated with a mental health problem.

Problem Statement

Mental health problems are a major public health concern and may be associated

with substance use (Conway et al., 2018). Individuals who have inadequate social support

over time are at a greater risk of experiencing mental health problems compared to

individuals who have a strong social support system (Holden et al., 2015). Substance use

issues may intensify with inadequate social support (Milner et al., 2016). A limited

number of studies have examined the quality of social support among individuals with

mental health problems (Amosu et al., 2016; Henchoz et al., 2016; Wang, Davis,

Wootton, Mottershaw, & Haworth, 2017) and the co-occurrence between substance use

and mental health problems (Conway et al., 2018).

Demographic differences such as race may play a role in substance use and mental

health problems and may contribute to young adult access to health care services

(Conway et al., 2018). While the occurrence of mental health problems, substance use,

and social support are well studied in the adult population, little is known about the

degree to which social support impacts mental health problems and substance use or

abuse among teens in MS and HS age groups. Therefore, further research is needed to

examine whether mental health problems are associated with social support and substance

abuse in MS and HS aged children as well as whether the problem increases with age.

This study examined retrospective data from a state-wide survey of youth in

MS and HS for substance use and social support associated with mental health problems.

Purpose of the Study

Using secondary data from the MIYHS 2017, the purposes of my study were to

determine (a) the relationship between mental health problems and substance use among

youth who are MS or HS students, (b) the relationship between mental health problems

and social support in youth who are MS or HS students, and (c) the relationship between

substance use and social support in youth who are MS or HS students. Data from the MS

and HS MIYHS 2017 were obtained from the Maine Center for Disease Control and

Prevention (Maine CDC) for the study. The MIYHS collected the data using specific

surveys developed for their respective age groups (MIYHS, 2017). The results of this

current secondary analysis may help practitioners to develop focused age-specific

interventions for students at risk of mental health conditions secondary to drug use and to

plan prevention programs for school-aged youths.

Research Questions and Hypotheses

This dissertation addressed the following research questions:

•Research Question 1: Using the MIYHS dataset, what is the relationship between

mental health problems and substance use among youth who are MS or HS

students?

H01: There is no relationship between mental health problems and substance use

among youth who are MS or HS students.

Ha1: There is a relationship between mental health problems and substance use

among youth who are MS or HS students.

•Research Question 2: Using the MIYHS dataset, what is the relationship between

mental health problems and social support in youth who are MS or HS students?

H02: There is no relationship between mental health problems and social support

among youth who are MS or HS students.

Ha2: There is a relationship between mental health problems and social support

among MS or HS students.

•Research Question 3: What is the relationship between substance use and social

support in youth who are MS or HS students?

H03: There is no relationship between substance use and social support in youth

who are MS or HS students.

Ha3: There is a relationship between substance use and social support in youth

who are MS or HS students.

Theoretical Framework

Two theoretical frameworks were used in this study: Relation regulation theory

(RRT) and the self-medication theory. The RRT provides a conceptual framework that

helps in assessing the direct impact of social support on mental health outcomes (Lakey

& Orechek, 2011). This theory maintains that supportive interaction has a positive impact

on mental health outcomes (John & Louise, 2013). Most importantly, the RRT holds that

individuals improve their mental health through the diversity of relationships (John &

Louise, 2013). For instance, if youth defined communication, interaction, and relationship

as a way of gaining social support, their behaviors and beliefs about social support is that

supportive interaction has a positive impact on mental health (John &

Louise, 2013).

The second theoretical framework is the self-medication theory. This theory helps

to conceptualize the association between substance use and mental health problems

(Lerissa et al., 2017) and to better understand substance use associated with mental health

problems among youth (Khantzian, 2017). Self-medication theory was developed to

address the concern about substance use by looking at the psychological underpinning of

substance use (Khantzian, 2017). For instance, the self-medication theory identifies and

addresses psychobiological factors that may cause substance use problems (Khantzian,

2017). This theory helps in identifying the reasons individuals use substances and may

help address the associated between substance use and mental health problems

(Khantzian, 2017). More detail on the self-medication and RRT frameworks is presented

in Chapter 2.

Nature of the Study

This study was a cross-sectional, retrospective, correlational design using a

quantitative approach. For this study, I used data from the 2017 Maine Integrated Youth

MIYHS. In this study, the definitions used by the MIHYS were adopted and include

•mental health problem, measured as a sad feeling or suicidal ideation for the

past 2 weeks (MIYHS, 2017); and

•social support, measured by love and support from family as well as

communication with family and friends.

These are captured by the following two questions: (a) “I have a family that gives me

love and support” and (b) “I have parents who are good at talking with me about things”

(MIYHS, 2017). Substance use was measured using the following question: “During your

life, how many times have you used any form of cocaine, including powder, crack, or

freebase?” (MIYHS, 2017). The independent variables include social support, substance

use, and age group. The dependent variable is mental health problems. Other variables

that were included in the study are county and number of individuals in the household

and parent’s social, economic status. The age group categorized MS or HS years of age.

In my study, I used secondary data from the 2017 MIYHS to examine mental health

problems associated with substance use and social support among youths. The data were

analyzed through logistic regression in which mental health problems is categorized as

yes or no.

The study focused on the analysis of the MS and HS student data from the

MIYHS. The sampling frame includes all public and quasi-public schools in the state of

Maine that enrolled at least 10 students with kindergarten-third grade, fifth-sixth grade,

seventh-eight grade (MS), and ninth through 12th grade (HS).

Definitions

Mental health problems: Manwell et al. (2015) defined mental health problems as

the absence of mental disease, which includes the biological, psychological, and social

factors that could affect a person's mental state and their function in society. The World

Health Organization (WHO, 2014) defined mental health as an individual’s ability to

realize their potential to cope with normal life events and able to contribute to society.

Dependent variable for assessing mental health problems was sad feeling or suicidal

ideation for past two weeks (MIYHS, 2017). Participants who are reporting sad feelings

or suicidal ideation were coded as 1 and 0 otherwise.

Social support: A formal or informal relationship that an individual has with

another person (Joni & Leonard, 2013). Social support that comes from family members,

friends, and peers is considered informal social support, whereas when it comes from

organization and healthcare professionals, it is considered formal social support (Joni &

Leornard, 2013). It is when an individual has friends, family, peers, and others to reach

out to when in need to help the individual focus on a positive outlook (University of

Minnesota, 2016). Social support can enhance the quality of an individual’s life by

providing a buffer in times of crisis (University of Minnesota, 2016). The understanding

of social support is that the individual feels accepted, receives care, and assistance during

difficult times (Joni & Leonard, 2013). Free sharing of resources to help improve

wellbeing and to allow interaction between individuals is also a form of social support

(Joni & Leornard, 2013). Social support was measured by love and support from family

as well as communication with family and friends. Independent variables were: (a) I have

a family that gives me love and support, and (b) I have parents who are good at talking

with me about things (MIYHS, 2017). Participants who reported having a family that

gives them love and support or having parents who are good at talking with them about

things were coded as 1 and 0 otherwise.

Substance use: The use of legal or illegal substances without a prescription

(Birkeland et al., 2017). The independent variables of substance use were measured by

this question: “During your life, how many times have you used any form of cocaine,

including powder, crack, or freebase?” (MIYHS, 2017). Response participants who

reported using methamphetamines or ecstasy during their life was coded as 1 and 0

otherwise.

Secondary data: Data derived from primary data, which is used to answer

research questions that the data were not developed to answer (Shamblen & Dwivedi,

2010).

Middle school (MS): MS refers to youth who are in Grades 7 and 8.

High school (HS): HS refers to youth who are in Grades 9-12.

Assumptions

Assumptions provide a structure about the belief of the methodology,

understanding, and the reality of research (Russel, 1995). An assumption made in this

study is that the participants provided an honest and truthful response to the MIYHS

survey and that the data was collected correctly. The data source is reliable. An

assumption underlying one of the hypotheses is that a lack of social support among

individuals negatively impacts mental health, particularly suicidal ideation and depressive

symptoms. Individuals who keep to themselves are more susceptible to suicidal ideation

and depression than are individuals who interact with others within a positive social

circle.

Scope and Delimitations

The research study used a quantitative, cross-sectional, correlational design with

secondary data from the MIYHS to examine mental health problems associated with

substance use and social support among youth. The MIYHS was conducted by the Maine

Department of Education who collaborated with the Maine CDC and Substance Abuse

and Mental Health Services in the Department of Health and Human Services (MIYHS,

2017). However, the survey is administered by the Pan Atlantic SMS Group administered

to youths who are in school. The survey was conducted in 2015 to assess the risk

behaviors of the Maine’s youth (MIYHS, 2017).

Secondary data are collected by someone else, such as a government agency,

schools, or recordkeeping organizations (Steve, 2017). The use of secondary data is less

time consuming and more cost-effective (Steve, 2017). Secondary data can be easily

accessed through existing data sources (Steve, 2017). Even though the study uses the

public and quasi-public schools in the state of Maine that enrolled at least 10 students,

there are some limitations because students who are homeschooled would not be

captured. The demographics of youth in Maine might not represent the demographics of

all youth in the United States. Therefore, the result might not be generalizable to the

United States population. This study focused on middle and high school students because

this is the age at which they are transitioning to young adults. The association of mental

health problems with substance use and social support were only considered at the public

school of Maine. Some of the population of the public school of Maine was excluded due

to grades. The excluded grades were grades below kindergarten-third grade. The

theoretical framework for understanding the association between mental health, social

support and substance use in excluded population is the social ecological model of human

development. This model is conceived as a set of nested structure for understanding

factors (social, emotional, etc.) that influence individuals’ development at varying degree

(Trach, Lee, & Hymel, 2018).

Limitations

This study was a descriptive, correlational study that assessed association between

mental health and substance use and social support in youths. The identification of

association does not mean there is causation between the variables (Seema, 2018). This is

a limitation because there might be other possible explanations for any observed

association. For example, an unknown genetic factor may be associated with mental

health, and the same genetic factor may associate with substance use (Adibsereshki,

Abdollahzadeh, Hassanzadeh, & Tahan, 2018). Therefore, the observed association might

be due to an unknown confounding factor. To address the limitation of confounding

factors, for this study, I controlled for covariates that may be associated with the key

variables of social support, substance abuse, and mental health problems. I controlled for

number of adults over 21 known to participants who in the past year have used marijuana,

crack, cocaine, or other drugs. The study results are from only Maine and may not be

generalizable to young adults in the whole United States. Some schools in Maine were

not able to participate in the survey due to grade reconfiguration or closing of the school

(Maine CDC, 2017). Therefore, the modules for the Grades K-3 and 5-6 were replaced by

an eligible school that may only closely match the schools that were ineligible for the

study. Therefore, an adequate and representative sample may not be available for the

study. Schools that failed to respond to the survey were also replaced by another school

that was chosen randomly to replace those nonresponding schools.

The MIYHS has a limitation because the study only focused on youth in school

(Maine CDC, 2017). Therefore, youths who are dropouts, absent, home-schoolers,

homeless, and runaway youth were missed or no response (Maine CDC, 2017). Some

schools had a small number of students who were enrolled in Maine’s island or rural

areas which led to non-response to survey due to exclusion (Maine CDC, 2017). The

language barrier is a limitation because the questionnaire is anonymous and was not

administered in multiple languages (Maine CDC, 2017). Therefore, individuals with

limited English language may not respond to the survey (Maine CDC, 2017). Another

limitation barrier is that teachers are not allowed to help students fill out the survey

stemming from confidentiality and to maintain consistency in the way the questions were

developed (Maine CDC, 2017). Therefore, students who are experiencing disability in

reading and cognitive impairment may not complete the survey or response to the survey

(Maine CDC, 2017). There may be bias due to self-reporting because the participants

may forget, deceive, or may not understand the questions, which may lead to

underreporting or overreporting (Maine CDC, 2017). Nonetheless, studies have shown

that surveys completed by young adults could provide valid and reliable information that

is needed for research (Maine CDC, 2017).

Significance of the Study

Identification of whether or not social support and substance use are associated

with mental health problems will help policymakers to develop better mental health

policies. The result of this analysis could help practitioners develop focused age-specific

interventions for students at risk of mental health conditions secondary to drug use and to

plan prevention programs for school-aged youth.

The self-medication theory may be used to identify strategies that will help

address the psychobiological factors that contribute to substance use problems (Tronnier,

2015). This study may generate data to better understand the nature of the

interrelationship between substance use, social support, and mental health problems (Hall

& Queener, 2007; Tronnier, 2015) and may help the practitioner develop an intervention

for youths experiencing substance use problems. The RRT may be used to understand an

individual’s social support perception about social support and its benefits to better health

outcomes (Woods Lakey, & Sain, 2016). By understanding social support and its

benefits, practitioners may develop ways to improve social support that will help buffer

stressful events; individuals may experience in their lives (Cohen & Will, 1985).

Therefore, the RRT may extend to practice to help youth access the social support needed

for their well-being. Evidence-based practice may help improve individual health

outcomes, so the study on mental health problems association with substance use and

social support may identify and buttress the need for intervention that can be used to

bring about positive social change in society. This may help an individual from

experiencing worsening of mental health problems that may lead to unproductive life in

society (Castello, 2016). In all, this study may contribute to the existing body of

knowledge on how the prevention of substance use and the promotion of social support

could lead to an improvement in mental health outcomes.

Significance to Theory

Khantzian et al. (2017) provided information on the self-medication theory that is

used for the psychological underpinnings of substance use that may lead to mental health

problems.

The concept of self-medication is used to identify ways to address

psychobiological factors that may cause substance use problems (Tronnier, 2015).

According to Tronnier (2015), self-medication theory provides a broader understanding

of substance use by individuals. Self-medication theory was used to explained substance

use association with mental health among youth (Hall & Queener, 2007). Woods et al.

(2016) used the RRT to predict the main effect of perceived support based on activity that

is shared. The RRT is used to understand better how social support is necessary to

prevent a mental health problem. The study examined mental health problems association

with substance use and social support among youth from the public and quasi-public

schools in the state of Maine but can be used in many facilities, also applying the

selfmedication theory and RRT in clinical practice.

The conceptual framework was used to strengthen the understanding of the

relationship between mental health problems, social support, and substance use in youth.

The theory used in this current study might not apply to other populations, such as young

adults or older adults, because the theory in this study was based on the youth population

only.

Significance to Practice

Public health professionals are in constant need of evolving empirical evidence to

inform evidence-based practice guidelines that will help improve patients’ health

outcomes. If an association between mental health and (a) social support and (b)

substance use among youth is identified, early intervention can be developed to improve

positive health outcomes. If an association is found to exist between mental health

problems and social support, it will provide data to show the importance of ensuring that

individuals with mental health problems have access to the social support that they need.

Moreover, any association found in this study will call attention to the merit of continued

sharing of data with practitioners who are in the best position to implement an

intervention that is needed to improve mental health outcomes.

Significance to Social Change

By evaluating the association between mental health problems, substance use, and

social support, specific and more focused intervention can be developed to promote

mental health outcomes. If these correlates of mental health problems are identified early

in young adults’ life, an intervention can be implemented early, and this will help youth

with mental health symptoms become more productive in society (Costello, 2016). If the

youth with mental health problems receive needed intervention early, this may help

worsening of the mental health problem, which could have an untoward effect on the

family, community, and population at large. Early intervention will not only help the

youth live a more productive life, but it will also help avert future negative effects of

mental health problems, including but not limited to the cost of hospitalization and the

youth inflicting harm on themselves. This will, without a doubt, bring about positive

social change in society (Costello, 2016). Early intervention in preventing substance use

among youth that may lead to mental health problems will increase positive social change

for the individual, community, and in the population (Spoth, Trudean, Redmond, & Shin,

2014).

Summary and Transition

Mental health problems continue to be a problem in society, and youth in the

public school in Maine experience the same problems (Milner et al., 2016). This study

focused on mental health problems associated with substance use and social support

among youths. There may be an increase in the number of youths experiencing mental

health problems associated with substance use and inadequate social support among

youths (Spoth et al., 2014). Even though there are many studies in other populations,

there is limited understanding when it comes to mental health problems associated with

substance use and social support among youths in public schools in Maine (Milner et al.,

2016).

Chapter 2 will present the review of the literature on mental health problems

association with substance use and social support.

Chapter 2: Literature Review

Using secondary data from the MIYHS 2017, the purpose of my study was to

determine (a) the relationship between mental health problems and substance use among

youth who are MS or HS students, (b) the relationship between mental health problems

and social support in youth who are MS or HS students, and (c) the relationship between

substance use and social support in youth who are MS or HS students. Existing literature

suggested that mental health problems are associated with substance use (Amosu et al.,

2016; Conway et al., 2018) and social support (Milner et al., 2016) in the adult or overall

population. However, the association between mental health problems and social support,

as well as between mental health problems and substance use varied between the overall

population and young adults (Milner et al., 2016).

Despite many studies on the association of social support and substance use with

mental health in the overall population, no research was identified showing that the

variables were associated among MS- and HS-aged children (Cheng et al., 2014; Jibeen,

2016; Lakey et al., 2016; Lerissa et al., 2017; Levula et al., 2016; Mason et al., 2014).

Therefore, the gap in knowledge regarding the association of social support and substance

use with mental health problems among young adults, including middle and high school-

aged children, was the focus of this current study.

In this literature review, I evaluated previous studies regarding mental health

problems and their association with substance use and social support. This chapter

presents the literature search strategy, literature review on mental health problems, social

support and substance use, theoretical framework, summary, and conclusion.

Literature Search Strategy

The databases used for the literature review include CINAHL Plus with full text,

Medline with full text, Proquest dissertations and theses, Proquest Nursing and Allied

health sources, Pubmed, CINAHL and Medline combined sources, Ovid full-text nursing

journals, and Psychology full-text sources. The key terms and phrases used for the

literature review included mental health problems, mental health, social support,

substance use, substance abuse, substance use/abuse intervention, MIYHS, suicidal

ideation in youths, sad feelings, and social support interventions.

The articles used for the literature review search included peer-reviewed articles

that describe mental health problems and their association with social support and

substance use in public and quasi-public schools and published from 2000 through 2018.

Much of the literature on the association of mental health problems with substance

use and social support focused on adults, college students, and prisoners, whereas a small

number of articles focused on mental health problems that may be associated with

substance use and social support in public schools (Amosu et al., 2016;

Cheng et al., 2014; Holden et al., 2015; Lerissa et al., 2017).

Theoretical Framework

Two theoretical frameworks were used to guide the study: the theory of

selfmedication and the RRT.

Theory of Self-Medication

Khantzian’s (2017) and David’s (1974a) theory of self-medication was used to

guide substance use treatment, so I used it in examining substance use in the current

study. Khantzian and David developed the theory to assess and treat drug addiction

problems. The self-medication was in the context of regulation where an individual’s

sense or understanding of regulation matches self-medication. When self-medication is

applied to substance use in public and quasi-public schools, the theory may be better

understood.

The self-medication theory proposed that there is an underlying reason for drug

use. Although Khantzian and Duncan are credited as the originators of the selfmedication

theory, Fenichel (1945) and Rado (1957) suggested that pressure was an underlying

motivation for drug use. Other authors, such as Glover (1956) and Ronsenfeld (1965),

have also contributed to this theory. The Khantzian model of self-medication presupposes

that drug addiction results from inadequate ability to control aggression; therefore, the

drug addict resorts to a drug as a way of controlling their aggressive drive, which leads to

physical dependency (Khantzian, 1997b, 1999). This self-medication hypothesis, which

originally included cocaine, was subsequently expanded to alcoholism and later

developed into a theory (Khantzian, 1997a, 1999). The fully developed selfmedication

hypothesis asserts that drug addiction results from venerable self-medication (Achalu,

2002).

Duncan’s (1974a) version of self-medication, rooted in behavioral theory, makes a

distinction between drug use and drug abuse. Duncan asserted that many of the illegal

drug users do not qualify as a substance abuser and that the majority of drug users have

control over their drug use. Duncan’s model is concerned with drug use without a

prescription, which could expose the individual to health risk. Duncan maintained that

drug use results from the pleasurable effect of the drug. Duncan’s model attempted to

explain drug dependence, a term used to replace the term addiction. The theory will help

explain the reason youths are using a drug that can lead to mental health problems, which

is a health risk.

Theory of Relational Regulation

Lazarus and Folkman’s (1984) RRT is used to explain how social support helps

individuals manage stress that they may be experiencing in their lives. However, enough

support is needed to help the individual cope with the stressor. The stress and coping

version of the RRT can explain stress-buffering effects but not the main effect of stress.

Therefore, the main effects of perceived social support associated with mental health

problems are likely to increase and multiply at a steadier pace than do the buffering

effects, so the stress and coping theory are not able to appropriately explain the main

effects of perceived social support associated with mental health problems (Lazarus &

Folkman, 1984).

The RRT maintained that individuals who think they have someone to assist them

in their times of need have better mental health outcomes than those who perceived low

social support. The purpose of RRT was to explain the main effect of perceived support

on mental health because this effect could not be explained by stress and coping theory.

The rationale in the use of RRT is that the association between social support and mental

health originates from people regulating their emotions through shared activity and

conversation. RRT has been found to have both buffering and a direct effect on mental

health. The regulation of emotion is relational because personal taste determines the topic

of conversion, activities, and providers of support (Lazarus & Folkman, 1984).

Cohen and Wills (1985) differentiated between stress-buffering and the main

effect, which is the foundation of the RRT. Stress buffering happens when an individual

has someone to protect them from the stress they are experiencing. Therefore, the main

effects happen when individuals have high social support with better mental health

outcomes, while other individuals have low social support, even with or without stress.

The stress-buffering theory, a version of the RRT, was developed to explain the social

support associated with mental health, which reflects stress buffering. Many authors,

including Barrera (1986), Cohen and Wills, Cutrona and Russell (1990), and Thoits

(1986), contributed to the development of the stress-buffering theory.

This stress-buffering theory is an extended version of the stress and coping theory,

developed by Lazarus (1996) and Lazarus and Folkman (1984). The stressbuffering

theory may be summarized into five different factors, such as life events, increased events

risk, the stability of social support, availability of social support, and how effective the

social support was able to relieve the stress. Even though the stressbuffering theory has

been used in many areas of research, there are empirical limitations because this theory

only evaluates inconsistency. However, Brown and Harris (1978) found consistency in

the stress-buffering theory. Figure 1 illustrates the RRT; the rationale in the use of RRT is

that the association between social support and mental health originates from people

regulating their emotions through shared activity and conversation. RRT has been found

to have both buffering and a direct effect on mental health. The regulation of emotion is

relational because personal taste determines the topic of conversion, activities, and

providers of support (Lazarus & Folkman, 1984).

Figure 1. Schematic representation of the relational regulation theory (RRT). Designed

for social support research.

Literature Review Related to Key Variables and Concepts

The literature review results are organized by the key variables and concepts:

mental health; social support; substance use; MIYHS; the association between mental

health and substance abuse; the association between mental health and social support; the

association between substance use and social support; existing approaches in the study of

mental health, substance use, and social support; theoretical framework; limitation of

previous approaches; the rationale for variables in this study; and the gap.

Mental Health

Mental health problems are a significant public health concern and are considered

to be a huge disease burden among young people in society (Cheng et al., 2014; Manwell

et al., 2015). As noted by Cheng et al. (2014), mental health is an individual’s state of

well-being, where these individuals understand their potential, can cope with daily stress

in their daily life, be productive, and contribute to their community. Mental health

problems in adolescents are considered to be strongly related to mental health in

adulthood.

These health problems can be influenced by biological and psychosocial factors

(Cheng et al., 2014; WHO, 2014). Other risk factors that affect mental health include

genetic, cognitive, temperamental, interpersonal, and family environment (Cheng et al.,

2014). In young adults, mental health problems impact educational achievement,

substance use, violence tendency, reproduction, and sexual health (Cheng et al., 2014).

Therefore, it is necessary to study the risk and the protective factors for mental health

problems among adolescents; such a study will help develop effective prevention

strategies (Cheng et al., 2014; Conway et al., 2018).

Social Support

Social support is an essential factor that helps the individual cope with changes in

their lives (Wang et al., 2017). Social support may help to alleviate distress and provide

hope in a time of distress (Cheng et al., 2014). Lack of social support can negatively

impact children and young adults’ life, especially those exposed to stressful life events

(Cheng et al., 2014). Providing social support in the family and the community may help

decrease distress and promote hope. Social support is crucial in early adulthood (Holden

et al., 2015). Individuals who lose their social networks are at risk for illness (Levula &

Harre, 2016).

Social support differs not only by gender but also by age. Women are more likely

than men to give and receive social support, to have more social friends, and to focus on a

social network (Milner et al., 2016). Relying on social group changes as individuals get

older. For example, adolescents have more social networks than older people (Milner et

al., 2016).

Substance Use

Substance use is a problem worldwide, and it is a risk factor for many health

problems. For example, cannabis is used illegally worldwide at a rate of 3.9% among

individuals aged 15-64, and it increases school dropout and motor vehicle accidents

(Henchoz et al., 2016).

Research shows that young adulthood is a crucial stage for substance use (Mason

et al., 2014), and young adults are more likely to drink heavier than older adults (Mason

et al., 2014).

The negative impact of substance use includes an increase in the cost of

healthcare, crime, and the yearly loss of productivity (Lerissa et al., 2017). Not only are

the financial costs a problem, but more importantly, substance use may predict health risk

behaviors that can cause long term negative health outcomes that include decreased

mental health (Amosu et al., 2016). Individuals who use substance sometimes face

memory problems, clear thinking and decreased attentiveness (Lerissa et al., 2017) which

affects youths, particularly at a vulnerable time because of neurodevelopment. Substance

use may affect the process peak and the function of neural development that alters the

youth's academic and social functioning (Henchoz et al., 2016).

Preventative measures are necessary to decrease health risks and prevent the

progression of multiple substance use (Henchoz et al., 2016). There is a need to

implement programs and policies that discourage early substance use.

Maine Integrated Youth Health Survey

The MIYHS was developed in 2009 to make an improvement in data quality,

decrease the financial burden for Maine schools and expenses (MIYHS, 2017). MIYHS

allows multiple survey requests while providing many health topics about youth

(MIYHS, 2017). The survey was administered in February 2015 (MIYHS, 2017).

The MIYHS was created and implemented in 2015 through collaboration between

the Maine Department of Education and the Maine department of health and human

services (MIYHS, 2017). The Pan Atlantic Research of Portland, Maine distributes the

2015 Youth Health Survey (MIYHS, 2017). Four surveys were composed and given to

students according to grades (MIYHS, 2017). The middle school and high school survey

collected data on many topics including mental health, social support and substance use

(MIYHS, 2017). The school must meet specific criteria in order to be eligible to

participate in the survey (MIYHS, 2017).

Association Between Mental Health Problems and Substance Use

Mental health and substance use are problematic because they are associated with

poor health outcomes, and about 20% of children and adolescents experience mental

health problems such as depression because of substance use (Cheng et al., 2014; Lerissa

et al., 2017). The use of a substance can adversely affect an individual’s mental health;

these health consequences from substance use extend to the increase of anxiety (Amosu

et al., 2016). An individual that is experiencing mental health problems along with

substance use is considered to be a co-occurring disorder. Depression is a major

contribution to substance use in individuals. Many individuals use substance use as a way

to cope with mental health (Lerrisa et al., 2017). Excessive substance use, such as

alcohol, may be associated with a mental health problem such as suicide, and young

adults with depression have a higher risk of comorbidity with suicidal ideation, anxiety,

and substance use (Mason et al., 2014).

Almost 20% of individuals with a mental health problem develop a substance use

problem in their lifetime, and about 55% of the individuals with mental health and

substance use get no treatment, while about 7.4% with both disorders received treatment

(Lerrisa et al., 2017). Some barriers prevent help-seeking for mental health and substance

use problems such as the stigma of mental health problems, lack of understanding about

mental illness in parents and young people, cultural competency, the financial cost, while

many individuals with mental health, alcohol, and substance use problems do not seek

help (McCann, Mugavin, Renzaho, & Lubman, 2016).

Association Between Mental Health Problems and Social Support

Worldwide, individuals are at higher risk for mental health problems that may be

associated with social support (Cheng et al., 2014). Social support contributed to the

mental health of individuals. Social support throughout young adult life may contribute to

psychosocial wellbeing in individual adult life (Holden et al., 2015). Social support

decreased the mortality rate and increased better mental health outcomes in individuals,

and early social support may help prevent mental health in adulthood (Holden et al.,

2015). Positive social support associated with mental health varies with gender financial

status and an individual stage in their life. However, people with poor social support may

experience poor mental outcomes and well-being (Holden et al., 2015). Social support

and mental health vary with gender, socioeconomic status, and the stage of life (Holden et

al., 2015). Poor social support and mental health may affect individual well-being as the

individual ages, and a decrease in social support affect one's mental health, while high

social support provides better mental health outcomes (Holden et al., 2015; Shahdad et

al., 2017). Social support can minimize daily stressor, and this can protect and improved

mental health outcomes (Shahdad et al., 2017).

Association Between Substance Use and Social Support

Substance use and social support affect an individual’s well-being. Social support

helps decrease the rate of substance use (Holden et al., 2015; Studer et al., 2017). Poor

social support from spouse, family, and friends put an individual at risk for substance use.

However, the benefits of social support depend on the type of support the individual

received (Studer et al., 2017). College students with no positive social networks are at

high risk for using substance, including alcohol, tobacco, and marijuana, while support

from friends who use substances may increase an individual’s risk, support from spouse

and family may be protective to prevent substance use ((Mason et al., 2014; Studer et al.,

2017).

Existing Approaches in Study of Mental health, Substance Use, and Social Support

Research Approach

Milner et al. (2016) assessed the association between social support and mental

health by age and gender using 13 years of cohort data. The authors used mental health

Inventory-5 (MHI-5). A social support scale was used which composed of 10 items. A

change in mental health was assessed using longitudinal fixed effects regression model

controlling for time. Results suggest that a higher level of social support helps improve

mental health. A cohort study using risk factors (C-SURF) and longitudinal study design

to assess risk and protective factors for substance use in adulthood (Studer et al., 2017).

The perception of social support was examined using a multidimensional scale (Jibeen,

2016). The MSPSS that consist of 12-items that got a score on a 7 point li kart-type scale

ranging from 1 through 7, where one strongly agrees, and seven is strongly disagreed

(Jibeen, 2016). Results showed that the coefficient of correlation that mental health

problems increased with the low level of familiar social support (Jibeen, 2016). The

authors assessed the differences in mental health and substance use among gender and

education levels using 634 American Indians from the Eastern Tribe (Ni et al., 2017). The

study analysis used logistic regression analysis of variance. The study showed that men

self-rated better mental health but binge drank and used illicit drugs more frequently than

females. Individuals who had a post-high school degree had better mental health than

individuals who did not have a high school diploma. The increase in education decreased

the use of a substance such as cigarette smoking. Illicit drugs include marijuana, cocaine,

and LSD.

Theoretical Framework

The theories used for this study is the strengthening families program for youth

10-14 and life skills training, and the life skill training is based on social learning theory

(Spoth et al., 2014). Personal social network framework used to understand better

substance use intervention and their supportive and unsupportive network to assess

participant’s behavior and their recovery (Tracy, Munson, Peterson, & Floersch, 2010).

The study used the risk, and the protective factors for mental health problems and the

impact on cultural and geographical variations have on mental health outcomes on

adolescents to help develop effective prevention strategies, and psychosocial factors help

determine an individual’s well-being (Cheng et al., 2014; Levula et al., 2016). The study

assessed the association between mental health scores and social network factors on

mental health across different life stages (Levula et al., 2016).

Rationale for Variables in this Study

The youth health survey (MIYHS) was used to improve data quality, decrease

collection expenses, and decrease survey burden on Main schools (MIYHS, 2017). The

collaborative survey decreased the burden on schools by using several survey requests

and gives local and state-level data a broad range of health topics about youths (MIYHS,

2017).

The tool that was used to collect the self-reported data can use in schools to assess

risk behaviors and protective factors of youths in society among states and nationwide

level data (MIYHS, 2017). The variables can be used by districts to create policy change

in school, shows a connection between youth academic performance, health risk

behaviors, schools climate problems, and help in getting grants for school programs

(MIYHS, 2017). Longitudinal studies found that social support may predict the outcome

of mental health problems (Holden et al., 2015). Low friendship security is related to

poor mental health outcomes, which include depressive symptoms (Mason et al., 2014).

Gap

More research is needed to address mental health problems in young people living

in already challenge with competing health and social burden and should not be neglected

(Cheng et al., 2014). More research is needed for substance use and mental health for the

implementation of early intervention for better health outcomes for youths (Conway et

al., 2018). My study is needed to examine the social support network properties across

different stages of life and not just in isolation of one another, and the influence of peer

networks on mental health symptoms among college students has been less studied.

Therefore, further research on the development of interventions that target the social

context of health is needed (Levula & Harre, 2016; Mason et al., 2014).

Summary and Conclusions

Mental health problem has been shown to be associated with substance use and

with a lack of social support, but little research has examined the problem among the

youth. Therefore, there is a gap in the examination of social support and substance use

and their association with mental health problems among youths. The databases used to

find full-text sources on mental health, substance use and social support include:

CINAHL Plus with full text, Medline with full text, Proquest dissertations and theses,

Proquest Nursing and Allied health sources, Pubmed, CINAHL and Medline combined

sources, Ovid full-text nursing journals, and Psychology. I used secondary data and

quantitative methods to examine the association between substance use and social

support. Two conceptual frameworks, the self-medication theory and the RRT served as a

foundation for understanding mental health problems and their association with social

support and substance use.

Chapter 3 presents the research design and rationale, methodology, including

population, sampling procedures, archival data, instrument, reliability, and validity. The

analysis tool and the data analysis used to analyze the data are explained.

Chapter 3: Research Method

Using secondary data from the MIYHS 2017, the purpose of my study was to

determine (a) the relationship between mental health problems and substance use among

youth who are MS or HS students, (b) the relationship between mental health problems

and social support in youth who are MS or HS students, and (c) the relationship between

substance use and social support in youth who are MS or HS students.

Chapter 3 will present the research design and rationale. The methodology

includes population, sampling procedures, and archival data, as well as the chosen

instrument, reliability, and validity. The analysis tool and the techniques used to analyze

the data are explained.

Research Design and Rationale

The study was a cross-sectional, retrospective, correlational design. For

assessment of the association between mental health problems with social support or with

substance use, the independent variables are social support and substance use, whereas

the dependent variable is a composite of mental health problems (suicidal ideation or sad

feelings). For the assessment of the association between substance use and social support,

the independent variable is social support, and the dependent variable is substance use.

There was no time constraint except the time necessary to get the secondary data

from the Maine CDC. The results of this quantitative, correlational study may help

provide a better understanding of the relationship among mental health problems, social

support, and substance use, and the type and strength of the relationship. Hence, results

from this study may provide strategies that support youth in the community by reducing

the risk of mental illness and substance-abuse-related outcomes while promoting social

support for youth. The MIYHS (2017) data can be used to identify the strengths and

challenges facing young people in society.

Methodology

This study involved a secondary, quantitative correlational analysis of data from

the MIYHS to determine the association of variables related to mental health problems,

substance abuse and social support. The sampling frames for the MIYHS comprised

Maine’s public and quasi-public schools that enrolled at least 10 students. The

questionnaire was administered to all students in the school districts in the state. Students

were required to answer the questions after consent forms were signed by parents. To be

eligible to participate in the MIYHS study, the students had to be in kindergarten-third

grade, fifth-sixth grade, seventh-eighth grade (MS), and ninth through 12th grade (HS).

The MS and HS surveys included four versions each (i.e., Version A, Version B, Version

C, and Version D).

Population

The population used for this study was drawn from the MS and HS students’

responses to the questionnaire in the state of Maine .Although other grades were included

in the original study survey, they were not included in the current secondary data

analysis. Access to the MIYHS data set required the principal investigator’s signed

confidentiality agreement, the institutional review board (IRB) approval, and project data

supervisor approval. I contacted the organization and spoke with the representative who

granted access to the data.

Sampling Procedures

A stratified random sampling was used to conduct the original study and included

all Maine’s public schools, quasi-public schools, and private schools that received about

60% of public funds (MIYHS, 2017). Schools with fewer than 10 students were excluded,

as were schools with alternative education and youth development centers. Schools that

did not respond were replaced by schools that were randomly selected from the sample

interval (MIYHS, 2017).

Archival Data

To gain access to the MIYHS data set, I signed a confidentiality agreement from

the CDC. I also secured IRB approval before commencing the study. Because this study

uses secondary data analysis, no informed consent was needed for the current study.

However, authorization was needed from the Maine CDC to use the data set.

Instrumentation and Operationalization of Constructs

For this study, I used secondary data analysis with no direct interaction with

participants. The data were collected by the Maine CDC using the MIYHS, which was

developed and implemented in 2015 by the collaborative effort of the Maine Department

of Education and the Maine Department of Health and Human Services. Four different

questionnaires labeled Versions A, B, C, and D were used to obtain information from

participants. Many questions used in the MIYHS survey were taken from other national

surveys. Some of the questions were taken from the National Youth Risk Behavior

Survey (YRBS). The reliability and validity of the YRBS questionnaire were conducted

by the Centers for Disease Control and Prevention (CDC, 2004) using two test-retest

reliability studies. One study was done in 1992, and the second study was done in 2000

(CDC, 2004).

Kappa was equal to 61%-100%, with no statistically significant differences

between the prevalence estimates for the first and second times the questionnaire was

administered. Kappa is used to measure the validity and reliability of the instrument

(CDC, 2004). Ten items (14%) had both Kappa less than 61% and significantly different

time-1 and time-2 prevalence estimates, which indicates that the reliability of those

questionnaire items is questionable (CDC, 2004). The CDC asserts that the validity of

adult self-reporting of behaviors measured by YRBS is not threatened by cognitive and

situational factors (CDC, 2004). Responses from the 7th_grade students were less

consistent than those from students in higher grades, suggesting that the questionnaire is

best suited for students in Grade 8 or higher (CDC, 2004).

The MIYHS survey results were appropriate for my study because the variables

used in my study were available from the question asked in the MIYHS and the variables

provided the answers to the questions that were used in the study to assess the association

between mental health with social support and substance use. The construct assessing

mental health, social support, and substance were taken from the MIYHS questionnaire:

Mental Health

•Item 14: (Have you ever seriously thought about killing yourself).

•Item 18: (During the past 12 months, did you ever feel so sad or hopeless

almost every day for two weeks or more in a row that you stopped doing

some usual activities);

•Item 20: (During the past 12 months, did you ever seriously consider

attempting suicide).

Substance Abuse

•Item 32: (Have you ever used an electronic vapor product), and

•Item 45: (About how many adults over 21 have you known personally

who in the past year have used marijuana, crack, cocaine, or other drugs).

•Item 63: (During the past 30 days, how many times did you take a

prescription drug (such as Oxycodone, Percocet, Vicodin, Codeine,

Adderall, Ritalin, or Xanax) without a doctor’s prescription).

•Item 69: (During your life, how many times have you used any form of

cocaine, including powder, crack or freebase), and

•Item 85 (About how many adults over 21 have you known personally

who in the past year have used marijuana, crack, cocaine, or other drugs).

Social Support

•Item 107: (I have support from adults other than my parents),

•Item 108: (I have a family that gives me love and support) are included in

the construct for social support.

(See Appendix for the whole questions.)

Reliability and Validity

MIYHS is a reliable and a standard instrument developed and validated by the

Maine CDC. Many questions in the MIYHS survey were taken from other national

surveys. Some of the questions were taken from the YRBS. The reliability and validity of

the latter were performed by the CDC using two test-retest reliability studies, one in 1992

and a second one in 2000 (CDC, 2004). In the first study, results showed that about

threefourths of the items used in the questionnaire were rated as having substantial or

higher reliability (CDC, 2004). Kappa was equal to 61%-100%, with no statistically

significant differences between the prevalence estimates for the first and second times the

questionnaire was administered. The second study result in 2000 showed that about one

of five items (22%) in the questionnaire had significantly different prevalence estimates

for the first and second times that the questionnaire was administered (CDC, 2004). In

2003, the CDC conducted a review of existing empirical literature to assess cognitive and

situational factors that might affect the validity of adolescent self-reporting of behaviors

measured by the YRBS questionnaire (CDC, 2004). The CDC determined that although

self-reports of these types of behaviors are affected by both cognitive and situational

factors, these factors do not threaten the validity of self-reports of each type of behavior

equally (CDC, 2004).

Data Analysis Plan

Because the sample collected by the MIYHS was stratified by county, public

health district, or schools, all estimates were weighted to reflect the distribution of the

variables in the middle and high school student population.

In this study, I addressed three research questions:

Research Question 1: Using the MIYHS dataset, what is the relationship between

mental health problems and substance use in youth who are MS years of age and youth

who are HS years of age?

H01: There is no relationship between mental health problems and substance use in

youth who are MS years of age and youth who are HS years of age.

Ha1: There is a relationship between mental health problems and substance use in

youth who are MS years of age and youth who are HS years of age.

To answer question 1, I used odds ratios from a logistic regression model to

determine the relationship, direction, strength, and significance between mental health

problems variables and substance use. Mental health and substance were nominal

(dichotomous). When the dependent variable is dichotomous, logistic regression model,

which gives the odds ratio, allows for adjustment of potential confounders in

multivariable analysis.

Research Question 2: Using the MIYHS dataset, what is the relationship between

mental health problems and social support in youth who are MS years of age and youth

who are HS years of age?

H01: There is no relationship between mental health problems and social support in

youth who are MS years of age and those who are HS years of age.

Ha1: There is a relationship between mental health problems and social support in

youth who are MS years of age and those who are HS years of age.

To answer question 2, I used the odds ratio from a logistic regression model to

determine the relationship, direction, strength, and significance between the mental health

variable and substance use. Mental health problems and substance will be nominal

(dichotomous). The odds ratio was used for the reasons given above.

Research Question 3: What is the relationship between substance use and social

support in youth who are MS years of age and youth who are HS years of age?

H01: There is no relationship between substance use and social support in youth

who are MS years of age and youth who are HS years of age.

Ha1: There is a relationship between substance use and social support in youth who

are MS years of age and youth who are HS years of age.

To answer question 3, I used the odds ratio from a logistic regression model to

determine the relationship direction strength and significance between mental health

variables and substance use. Mental health and substance used nominal (dichotomous).

The justification for using the odds ratio is provided in question 1 above.

Threats to Validity

The original questionnaires were tested and retested and were found to be valid

and reliable. To ensure that the design features are maintained during analysis, all

estimates were weighted to reflect the distribution of the variables in the middle and high

school student population (CDC, 2004). In 2003, the CDC conducted a review of existing

empirical literature to assess cognitive and situational factors that might affect the validity

of adolescent self-reporting of health risk behaviors measured by the YRBS questionnaire

(CDC, 2004). The CDC determined that although self-reports, health risk behaviors are

affected by both cognitive and situational factors, these factors do not threaten the

validity of self-reports of each type of behavior equally (CDC, 2004). Further details can

be found in the CDC codebook.

Ethical Procedures

This current study included analysis of secondary data; therefore, no informed

consent was needed. Authorization for use of the MIYHS was obtained from the Maine

CDC after I signed a confidentiality agreement and the Walden IRB approval was

obtained. No data access or analysis began until both the CDC and the Walden IRB

approvals were obtained.

Summary

This chapter presents the design and rationale for the current study. Using the

MIYHS survey, I analyzed the MIYHS secondary data to assess whether an association

exists between mental the variables of health problems, substance use, and social support.

For the current study, I focused on MS and HS students’ data from Maine’s public

quasi-public, and private schools from the 2017 MIYHS. To ensure that the design

features (e.g., stratification, cluster, etc) were maintained during analysis, I used

weightings consistent with the original survey results reported by the CDC to reflect the

distribution of the variables in the middle and high school student population.

Multivariable logistic regression used to assess whether an association exists between

dependent and independent variables, and this regression analysis enabled adjustment for

possible confounders. Chapter 4 will provide the study results and summary.

Chapter 4: Results

Introduction

Using secondary data from the MIYHS 2017, the purpose of my study was to

determine (a) the relationship between mental health and substance use among youth who

are MS or HS students, (b) the relationship between mental health problems and social

support in youth who are MS or HS students, and (c) the relationship between substance

use and social support in youth who are MS or HS students.

I addressed the following research questions:

Research Question 1: Using the MIYHS dataset, what is the relationship between

mental health problems and substance use in youth who are MS years of age and youth

who are HS years of age?

H01: There is no relationship between mental health problems and substance use in

youth who are MS years of age and youth who are HS years of age.

Ha1: There is a relationship between mental health problems and substance use in

youth who are MS years of age and youth who are HS years of age.

Research Question 2: Using the MIYHS dataset, what is the relationship between

mental health problems and social support in youth who are MS years of age and youth

who are HS years of age?

H01: There is no relationship between mental health problems and social support in

youth who are MS years of age and those who are HS years of age.

Ha1: There is a relationship between mental health problems and social support in

youth who are MS years of age and those who are HS years of age.

Research Question 3: What is the relationship between substance use and social

support in youth who are MS years of age and youth who are HS years of age?

H01: There is no relationship between substance use and social support in youth

who are MS years of age and youth who are HS years of age.

Ha1: There is a relationship between substance use and social support in youth who

are MS years of age and youth who are HS years of age.

Mental health problems were measured based on MIYHS items that asked

questions on sad feelings or suicidal ideation for the past 2 weeks. The items with

questions on mental health problem are as follows:

•During the past 12 months, did you ever feel so sad or hopeless almost every

day for 2 weeks or more in a row that you stopped doing some usual

activities?

•During the past 12 months, did you ever seriously consider attempting

suicide?

Social support was measured based on MIYHS items that asked questions about love and

support from family as well as communication with family and friends. These were

captured by the following two items:

1. I have a family that gives me love and support.

2. I have parents who are good at talking with me about things.

Substance use was measured based on MIYHS items that asked questions about the use

of substances. The following item captured information on substance use:

• During your life, how many times have you used any form of cocaine,

including powder, crack, or freebase.

This chapter will present data collection, study results, and a summary.

Data Collection

The Maine CDC collected data from February 2nd through February 13th, 2015, for

MS and HS youths using specific surveys developed for their respective age groups.

The MIYHS was developed and implemented in 2015 by the collaborative effort of the

Maine Department of Education and the Maine Department of Health and Human

Services, and The Pan Atlantic Research of Portland, Maine administered the 2015 Youth

Health Survey (MIYHS, 2017). Many questions in the MIYHS survey were taken from

other national surveys such as the YRBS. The reliability and validity of the YRBS

questionnaire were evaluated by the CDC using two test-retest reliability studies (CDC,

2004). The MIYHS is appropriate for this study because the variables used in the survey

contained the variables that were used in the study to assess the association between

mental health problems with social support and substance use.

Sampling Process

The CDC used stratified random sampling to obtain participants. The sampling

frame included all of Maine’s public schools, quasi-public schools, and private schools

that had about 60% of public funds. Schools with fewer than 10 students were excluded,

as were schools with alternative education and youth development centers. The sample

interval of 14.33 was computed by dividing the frame of 344 participants with the desired

sample size of 24 schools. A random number was chosen from 1 to 14.33, and a random

number was 1. Therefore, the first school was chosen, followed by every 14.33th school

rounded to the nearest integer taken. Schools that did not respond were replaced by

schools that were randomly selected from the sample interval (MIHYS, 2017). During the

data collection process, some schools were no longer eligible to participate in the survey

due to the reconfiguration of the grades or school closure. Therefore, these schools were

replaced to ensure that there was an adequate sample size and response rate.

Weighting

For this study, the pre-computed base weight and the principal sampling unit

(PSU) provided by the CDC were used. The computational details of the weight are

presented in the MIYHS methodology book (CDC, 2015). A brief outline of the method

used by the CDC for computing the weight is as follows. For the MS and HS survey, all

the schools that were eligible received an initial base weight of one (1). For the school

that participated in the MS and HS surveys, the base weight was computed to reflect the

probability of the specific version assignment. The count from summarizing the response

data was used to obtain the version probability of version assignment. For the survey

version assigned to a class, the ratio of the class version to all the classes within the

school was used. To obtain the overall probability of each version assignment, the school

probability for that version was multiplied for that class version.

The base weight for the school and class within the school version assignment

process was calculated as one divided by the assignment probability of each version.

According to the Maine CDC (2017), the version base weights were then computed as 1

divided by the product of the stage weight, that is, Base weight = 1 / mod_pM, M Є (A,

B, C, D). The computed weight, principal sampling unit, and stratification variable for

different levels of analysis were provided in the data sets by the Maine CDC. Other

methodological details are found in the codebook (MIYHS, 2015).

To gain access to the MIHYS data set, I signed a confidentiality agreement. I also

signed a data sharing and protection agreement. I obtained IRB approval from Walden

University and the Maine CDC before I commenced with the data analysis. Because this

study is a secondary data analysis, no informed consent was needed from participants.

However, I obtained authorization from the Maine CDC to use the MIYHS dataset for

analysis. Once all documentation was completed, a USB drive with single-user

passwordprotected data was sent to me by the CDC Maine.

Change in Planned Analysis

The original goal, as documented in earlier chapters, was to measure substance

use using the following questions:

•During your life, how many times have you used methamphetamines (also

called speed, crystal, crack, or ice)?

•During your life, how many times have you used ecstasy (also called

MDMA)?

However, these two questions were not in the data set I received from Maine CDC.

Therefore, to offset this data limitation, I used the following substitute question for

substance use: During your life, how many times have you used any form of cocaine,

including powder, crack, or freebase?

Data for the Current Study

For the current study, I used the dataset provided by the CDC after selecting the

specific questions needed for my analysis.

Research Question 1: Using the MIYHS dataset, what is the relationship between mental

health and substance use in youth who are MS years of age and youth who are HS years?

H01: There is no relationship between mental health and substance use in youth

who are MS years of age and youth who are HS years.

Ha1: There is a relationship between mental health and substance use in youth who

are MS years of age and youth who are HS years.

To answer Research Question 1, I used:

•Item 14: Have you ever seriously thought about killing yourself?

•Item 18: During the past 12 months, did you ever feel so sad or hopeless

almost every day for two weeks or more in a row that you stopped doing

some usual activities;

•Item 20: During the past 12 months, did you ever seriously consider

attempting suicide.

•Item 32: Have you ever used an electronic vapor product?

•Item 69: During your life, how many times have you used any form of

cocaine, including powder, crack, or freebase.

Possible confounder

•Item 85: About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs?

•Item 45: About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs?

Research Question 2: Using the MIYHS dataset, what is the relationship between

mental health problems and social support in youth who are MS years of age and youth

who are HS years of age?

H01: There is no relationship between mental health problems and social support in

youth who are MS years of age and those who are HS years of age.

Ha1: There is a relationship between mental health problems and social support in

youth who are MS years of age and those who are HS years of age.

To answer Research Question 2, I used:

•Item 18: During the past 12 months, did you ever feel so sad or hopeless

almost every day for two weeks or more in a row that you stopped doing some

usual activities;

•Item 107: I have support from adults other than my parents.

•Item 108: I have a family that gives me love and support.

Possible confounder

•Item 45: About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs?

•Item 85: About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs?

Research Question 3: What is the relationship between substance use and social

support in youth who are MS years of age and youth who are HS years of age?

H01: There is no relationship between substance use and social support in youth

who are MS years of age and youth who are HS years of age.

Ha1: There is a relationship between substance use and social support in youth who

are MS years of age and youth who are HS years of age.

To answer research question three, I used:

•Item 63: During the past 30 days, how many times did you take a prescription

drug (such as Oxycodone, Percocet, Vicodin, Codeine, Adderall, Ritalin, or

Xanax) without a doctor’s prescription?

•Item 69: During your life, how many times have you used any form of

cocaine, including powder, crack, or freebase.

Possible confounder

•Item 45: About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs?

•Item 85: About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs?

•Item 108: I have a family that gives me love and support) are included in the

construct for social support.

Statistical Analysis

After importing the data into SPSS, I cleaned the data to get it ready for analysis.

Students with sixth grade, Ungraded or other grades, or no response to grade question (In

what grade are you?) have their grade set to missing prior to analysis (per communication

with Maine CDC). I recoded some multi-category variables into two categories (yes or

no) to avoid a sparse category that might cause convergence problems. The recoded

variables are presented in Table 1. Because the sample collected by the MIYHS was

stratified by county, public health district, or schools, I applied weight to all estimates to

reflect the distribution of the variables in the middle and high school student population

in Maine. Without weighting to reflect the survey design, any analysis will be invalid.

Table 1

Recoded Items in Current Study

Items Original data response Current study recoded

categories response categories

Item 69: During your life,

how many times have you

used any form of cocaine,

including powder, crack, or

freebase?

1 = A. 0 times

2 = B. 1 or 2 times

3 = C. 3 to 9 times

4 = D. 10 to 19 times

5 = E. 20 to 39 times

6 = F. 40 or more times,

1 is recoded to 0 (None) 2

to 6 is recoded to 1 (Yes:

at least once)

Missing is recoded to

missing

Possible Confounder:

Item 85 (for HS)/ Item 45 (for

MS): About how many adults

over 21 have you known

personally who in the past

year have used marijuana,

crack, cocaine or other

drugs)?

1 = A. None

2 = B. 1 adult

3 = C. 2 adults

4 = D. 3 or 4 adults

5 = E. 5 or more adults,

1 is recorded to 0 (none)

2 to 5 is recorded to 1

(Yes)

Missing is recorded to

missing

Item 107: I have support from

adults other than my parents

1 = A. Not at all or

rarely

2 = B. Somewhat or

sometimes

3 = C. Very or

often 4 = D. Extremely or

almost always.

1 is recorded to 0 (none)

2 to 4 is recorded to 1

(Yes)

Missing is recorded to

missing

Item 108: I have a family that

gives me love and support.

1 = A. Not at all or

rarely

2 = B. Somewhat or

sometimes

3 = C. Very or

often 4 = D. Extremely or

almost always,

1 is recorded to 0

(No/rarely)

2 to 4 is recorded to 1

(Yes: I receive love and

support)

Missing is recorded to

missing

Item 63: During the past 30

days, how many times did you

take a prescription drug (such

as Oxycodone,

Percocet, Vicodin, Codeine,

Adderall, Ritalin, or Xanax)

without a doctor’s

prescription?

1 = A. 0 times

2 = B. 1 or 2 times

3 = C. 3 to 9 times

4 = D. 10 to 19

times 5 = E. 20 to 39

times

6 = F. 40 or more times.

1 is recorded to 0

(None: 0 times)

2 to 6 is recorded to 1

(Yes: at least once) Missing

is recorded to missing

I selected questions that appear on all four questionnaire versions. Because I was

interested in the analysis at all levels, the weighting, principal sampling unit, and strata

reflected this level of analysis. The reason for analysis at all levels is to ensure results can

be applied to MS and HS students in the state of Maine. The weighting variables to

reflect different levels of analysis were provided by the CDC. To prepare the data for

state-level analysis, I created a dataset with appropriate variables that reflect the complex

survey design for this chosen level of analysis. These variables include str_notschool for

the strata, psu_notschool for principal sampling unit or cluster, and wstaabcd for the

weight; these variables were provided by CDC. I applied the finite population correction

(FPC) for estimating variance under simple random sampling assumption (FPC Factor,

2008).

Because of the very large sample size (e.g., N= 35,503 for HS), the p-value for

testing association had statistical significance even when the effect (odds ratio) size

lacked practical importance. Therefore, the confidence interval was considered more

appropriate for testing hypotheses since the significance test using p-value can be

inverted to produce a confidence interval (Knapp, 2017). This is consistent with the

efforts of American Statistical Associate in discouraging widespread wrongful use of

pvalues (Wasserstein & Lazar, 2016). Using univariate and multivariable logistic

regression, I calculated the odds ratio for assessing the association between mental health

problems and social support and that between mental health problems and substance use.

I also used logistic regression to compute odds ratios for assessing the association

between social support and substance use. A reason for the multivariate logistic

regression is to assess whether the added covariate (adult over 21 that use substances)

will be a confounder of the identified association from the univariate model. This

confounding will be assessed by change in the univariate odds ratio when the variable is

added into the multivariable logistics regression model. The multivariable logistic

regression model adjusted for a response to the following item: About how many adults

over 21 have you known personally who in the past year have used marijuana, crack,

cocaine, or other drugs?

An important assumption of logistic regression is that the log odds (logit) of the

dependent variable are linearly related to the independent variable. Although

HosmerLemeshow goodness of fit test is often used for assessing goodness-of-fit for

logistic regression, this test would be inappropriate in complex surveys as weighting

invalidate the assumption of independence, identical distribution required for the

HosmerLemeshow goodness of fit test (Archer, Lameshow, & Hosmer, 2007; Shah,

Institute, & Barnwell, 2003). A goodness of fit test was based on pseudo R-square.

Analyses were performed using SPSS Version 26 premium with an add-on for complex

surveys.

Results

Demographics

In total, there are 35,503 participants for HS. Weighted data available on HS were

55,088; of these, 26,456 are female, and 28296 were male. The majority (90.0%) are

white, and least (0.1%) represented are Native Hawaiian or other Pacific Islander. Table 2

present the demographic characteristics of the participants.

Table 2

Demographics

Characteristics

HS (High School) MS (Middle School)

Actual

count

Weighted

count

% 95% CI Actual

count

Weighted

count % 95% CI

Age (years)

10 or younger

12 or younger

18 or older

N/A

160

5193

9359

9185

7887

3524

N/A

190

109

7354

13532

13962

13502

6439

N/A

0.3

0.2

13.3

24.6

25.3

24.5

11.7

N/A

0.3, 0.4

0.1, 0.3

12.8, 13.9

24.0, 25.2

24.8, 25.9

23.8, 25.2

11.0, 12.4

5038

9059

4153

167

N/A

7384

13264

6105

235

N/A

0.1

0.3

27.2

48.9

22.5

0.9

0.1

N/A

0.0, 0.1

0.2, 0.4)

26.0, 28.5

48.0, 49.8

21.5, 23.5)

0.7, 1.1)

0.0, 0.1)

N/A

Sex

Female

Male

17742

17395

26456

28296

48.3

51.7

47.7, 49.0

51.0, 52.3

9158

9234

13071

13871

48.5

51.5

47.6, 49.4

50.6, 52.4

Grades

7th

8th

9th

10th

11th

12th

Ungraded or

other grade

N/A

9698

9465

8636

7100

149

N/A

13681

13873

13517

13386

200

N/A

25.0

25.4

24.7

24.5

0.4

N/A

24.3, 25.8

24.7, 26.0

24.1, 25.4

23.4, 25.6

0.3, 0.5

9220

9069

N/A

13508

13301

N/A

50.2

49.5

N/A

0.3

48.5, 52.0

47.7, 51.2

N/A

0.2, 0.4

Ethnicity

Hispanic or

Latino

Yes

1849

32706

1500

52282

2.8

97.2

2.6, 3.0

97.0, 97.4

958

16228

668

24483

2.7

97.3

2.4, 2.9

97.1, 97.6

Table 2 continued

Demographics

Characteristics

HS (High School) MS (Middle School)

Actual

count

Weighted

count

% 95% CI Actual

count

Weighted

count

% 95% CI

Race

American Indian

951

725

1.4

1.2, 1.5

576

402

1.6

1.4, 1.8

or Alaskan

Native

Asian

Black or African

American

Hispanic Native

Hawaiian or

Other Pacific

Islander

White

Multiple Races

1133

1086

1849

27644

1444

984

892

1500

47920

1158

1.8

1.7

2.8

0.1

90.0

2.2

1.3, 2.5

1.1, 2.5

2.6, 3.0

0.1, 0.2

88.8, 91.1

2.0, 2.4

351

648

958

13414

942

251

455

668

22346

659

1.0

1.8

2.7

0.2

90.0

2.7

0.8, 1.2

1.1, 3.0

2.4, 3.0

0.1, 0.2

88.6, 91.3

2.4, 2.9

In total, there were 18,706 participants for MS. Weighted data available on MS

were 27115; of these, 13071 are female, and 13871 are male. Similar to the data for HS,

the majority of the participants in MS (90.0%) are white, and least (0.2%) represented are

Native Hawaiian or other Pacific Islander.

Answers to Research Questions

Research Question 1. Using the MIYHS dataset, what is the relationship between

mental health and substance use in youth who are MS years of age and youth who are HS

years?

H01: There is no relationship between mental health and substance use in youth

who are MS years of age and youth who are HS years.

Ha1: There is a relationship between mental health and substance use in youth who

are MS years of age and youth who are HS years.

Since the 95% confidence interval for the odds exclude one (1), I rejected the null

and conclude that the association between mental health and substance use is statistically

significant.

To answer research question # 1, I conducted a statistical analysis using odds

ratios to determine the relationship between mental health (dependent variable) and

substance use (independent variable). The results are presented below.

High school: Frequencies. Of the 26,034 weighted participants for HS,

approximately (14.7%) of the participants indicated they seriously consider attempting

suicide during the past 12 months, while 85.3% indicated not seriously consider

attempting suicide. About four percent (4.3%) of the participants indicated that they had

used cocaine, including powder, crack, or freebase, whereas 95.7% have indicated no use

of cocaine, including powder, crack, or freebase. Table 3 presents the frequency of the

participants’ responses.

Table 3

Prevalence of Mental Health and Substance Use Among HS Students

Items Actual count Weighted count Percent: 95%

CI)

Ever seriously consider attempting

suicide during the past 12 month

Yes

5296

29416

3827

22207

14.7

85.3

Have you used any form of

cocaine, including powder, crack

or freebase

Yes

860

16130

1112

24922

4.3

95.7

High school: Univariate odds ratio. The odds of seriously consider attempting

suicide during the past 12 months among participants who used any form of cocaine,

including powder, crack or freebase is 3.94 (CI: 3.36,4.61) times that of the odds among

participants who do not use any form of cocaine, including powder, crack or freebase. In

other words, the use of any form of cocaine, including powder, crack, or freebase is

associated with 293% higher odds of seriously considering attempting suicide in the past

12 months. Since the 95% confidence interval for the odds excludes one (1), the odds are

statistically significant. Table 4 presents the univariate analysis of the participants. Table

4 presents the odds ratio.

High school: Multivariable or adjusted odds ratio. Controlling for adults over 21

around participants who have used illegal substances in the past for HS, the odds of ever

seriously considering attempting suicide during the past twelve months among

participants who indicated the used of cocaine, including powder, crack or freebase is

3.057 increased in odds than among participants who indicated no use of cocaine,

including powder, crack or freebase. This suggests that the use of cocaine, including

powder, crack, or freebase, is associated with 205.7% higher odds of participants ever

seriously considering attempting suicide during the past twelve months. Table 4 presents

the adjusted odds ratio between mental health and substance use.

Table 4

Univariate and Multivariable Odds Ratio for Assessing the Association Between Mental

Health and Substance Use Among HS Students

Items

Univariate odds ratio ( 95%

CI)

Multivariable odds

ratio (95% CI)

Dependent: Ever seriously

consider attempting suicide during

the past 12 month

Have you used any form of

cocaine, including powder, crack

or freebase

Yes vs. No

3.935 (3.360, 4.608)

3.057 (2.383, 3.924)

Note. The multivariable logistic regression adjusted for the following question: About

how many adults over 21 have you known personally who in the past year have used

marijuana, crack, cocaine, or other drugs?

Middle school: Frequencies. Of the 12,961 weighted participants for MS,

approximately sixteen percent (16.5%) of the participants seriously thought about killing

them self, while 83.5% indicated no thought about killing them self. About ten percent

(10.3%) of the participants indicated ever used an electronic vapor product, whereas

89.7% have indicated no use of an electronic vapor product. The odds of participants who

thought about killing themselves is 3.692 (CI: 3.174, 4.294) times that of the odds among

MS participants who do not use an electronic vapor product. In other words, serious

thought about killing oneself in MS is associated with 269% higher odds of using an

electronic vapor product. Since the 95% confidence interval for the odds excludes one

(1), the odds are statistically significant. Table 5 presents the frequency of the MS

participants.

Table 5

Frequency of Prevalence of Mental Health and Substance use Among MS Students

Items

Actual count

Weighted

count

Percent: 95%

Have you ever seriously thought about

killing yourself

Yes

3074

15155

2142

10820

16.5

83.5

Have you ever used an electronic vapor

product Yes

No 991

8072 1339

11622

10.3

89.7

Middle school: Multivariable or adjusted odds ratio. Have you ever seriously

thought about killing yourself (dependent variable). Have you ever used an electronic

vapor product (independent variable). The model did not converge for the MS. Table 6

presents the odds ratio and adjusted odds ratio.

Table 6

Univariate and Multivariable Odds Ratio for Assessing the Association Between Mental

Health and Substance Use Among MS Students

how many adults over 21 have you known personally who in the past year have used

marijuana, crack, cocaine, or other drugs?

Items

Univariate odds ratio (

95% CI)

Multivariable odds

ratio (95% CI)

Dependent: Have you ever

seriously thought about killing

yourself

Have you ever used an

electronic vapor product

Yes vs. No

3.692 (3.174, 4.294)

Model did not

converge

Note. The multivariable logistic regression adjusted for the following question: About

Research Question 2. Using the MIYHS dataset, what is the relationship between

mental health problems and social support in youth who are MS years of age and youth

who are HS years of age?

To answer research question # 2, I conducted a statistical analysis using the odds

ratios to determine the relationship between mental health problems (dependent variable)

and social support (independent variable).

High school: Frequencies. Of the 34,989 weighted participants for HS,

approximately twenty-six percent (26.9%) of the participants indicated feeling sad or

hopeless almost every day for 2 weeks or more in the past 12 months, while 73.1%

indicated not feeling of sad or hopeless for the same time period. About ninety-one

percent (91.9%) of the participants indicated that they have social support, while 8.1%

indicated no social support. Table 7 presents the frequency of the participants.

Table 7

Frequency of Prevalence of Mental Health and Social Support Among HS Students

Items

Actual count Weighted count Percent: 95%

Feeling sad or hopeless almost

every day for two weeks or more

in the past 12 months

Yes

9462

24835

9417

25571

26.9

73.1

Support from adults other than

my parents

Yes

20985

2070

32154

2834

91.9

8.1

High school: Univariate odds ratio. The odds of feeling sad or hopeless almost

every day for 2 weeks or more in the past 12 months among participants who have

support from adults other than parents is 0.467 (CI: 0.419,0.521) times that of the odds

among participants without support from adults other than parents. In other words,

support from adults other than parents is associated with 53% lower odds of feeling sad

or hopeless almost every day for 2 weeks or more in the past 12 months. Since the 95%

confidence interval for the odds excludes one (1), the odds are statistically significant.

Table 8 presents the odds ratio.

High school: Multivariable or adjusted odds ratio. Controlling for adults over 21

around participants who have used substance in the past for HS, the odds of feeling sad or

hopeless almost every day for 2 weeks or more during the past twelve months among

participants who indicated support from adults other than parents is 0.422 times the odds

among participants who indicate no support from adults other than parents. This suggests

that support from adults other than parents is associated with 57.8% lower odds of feeling

sad or hopeless almost every day for 2 weeks or more. Table 8 presents the adjusted odds

ratio.

Table 8

Univariate and Multivariable Odds Ratio for Assessing the Association Between

Mental Health and Social Support Among HS Students

how many adults over 21 have you known personally who in the past year have used

marijuana, crack, cocaine, or other drugs?

Middle school: Frequencies. Of the 6,442 weighted participants of MS,

approximately nineteen percent (19.4%) of the participants indicate that they felt so sad

or hopeless almost every day for 2 weeks or more in a row that they stopped doing some

unusual activities, while 80.6% indicated not feeling so sad or hopeless almost every day

for 2 weeks or more in a row that they stopped doing some unusual activities. Most

(97.4%) of the participants indicated that the family gives them love and support, whereas

Items

Univariate odds ratio ( 95%

CI)

Multivariable odd ratio

(95% CI)

Dependent: Feeling sad or

hopeless almost every day

for 2 weeks or more in the

past 12 months

Support from adults

other than my parents

Yes vs. No

0.467 (0.419, 0.521)

0.422 (0.348, 0.511)

Note. The multivariable logistic regression adjusted for the following question: About

2.6% have indicated that no family gives them love and support. Table 9 presents the

frequency of the participants.

Table 9

Frequency of Prevalence of Mental Health and Social Support Among MS Students

Items

Actual count Weighted count Percent: 95%

Have you ever felt so sad or

hopeless almost every day for 2

weeks or more in a row that you

stopped doing some unusual

activities

Yes

2036

7024

1247

5195

19.4

80.6

Family give me love and support

Yes

12458

435

6273

169

97.4

2.6

Middle School: Univariate Odds Ratio. The odds of feeling so sad or hopeless

almost every day for 2 weeks or more in a row that participants stopped doing some

unusual activities is 0.249 (CI: 0.170,0.364) times that of the odds among participants

who had family that give them love and support. In other words, family love and support

was associated with 75% lower odds of feeling so sad or hopeless almost every day for 2

weeks or more in a row. Since the 95% confidence interval for the odds excludes one (1),

the odds are statistically significant. Table 10 presents the odds ratio.

Middle school: Multivariable or adjusted odds ratio. Controlling for adults over

21 around participants who had used substance in the past for MS, the odds of feeling sad

or hopeless almost every day for 2 weeks or more among participants who indicated their

family gives them love and support was 0.285 times the odds among participants who

indicate their family did not give them love and support. This suggests that family love

and support are associated with 71.5% lower odds of feeling sad or hopeless almost every

day for 2 weeks or more. Table 10 presents the adjusted odds ratio.

Table 10

Univariate and Multivariable Odds Ratio for Assessing Association Between Mental

Health and Social Support Among MS Students

how many adults over 21 have you known personally who in the past year have used

marijuana, crack, cocaine, or other drugs?

Items

Univariate odds ratio ( 95%

CI)

Multivariable odd ratio

(95% CI)

Dependent: Have you ever

felt so sad or hopeless

almost every day for 2

weeks or more in a row

that you stopped doing

some unusual activities

Family give me love

and support Yes vs. No

0.249 (0.170, 0.364)

0.285 (0.191, 0.425)

Note. The multivariable logistic regression adjusted for the following question: About

Research Question 3. What is the relationship between substance use and social

support in youth who are MS years of age and youth who are HS years of age?

H01: There is no relationship between substance use and social support in youth

who are MS years of age and youth who are HS years of age.

Ha1: There is a relationship between substance use and social support in youth who

are MS years of age and youth who are HS years of age.

Since the 95% confidence interval for the odds exclude one (1), I rejected the null

and conclude that the association between substance use and social support is statistically

significant.

To answer Research Question 3, I conducted a statistical analysis using the odds

ratios to determine the relationship between mental health (dependent variable) and social

support (independent variable).

High school: Frequencies. Of the 24,779 weighted participants of HS,

approximately four percent (4.3%) of the participants indicate using any form of cocaine,

including powder, crack, or freebase, while 95.7% indicated not using any form of

cocaine, including powder, crack or freebase. About ninety-four percent (94.3%) of the

participants indicated family gives them love and support, whereas 5.7% have indicated

not having family give them love and support. Table 11 presents the frequency of the

participants.

Table 11

Frequency of Prevalence of Substance Use and Social Support Among HS Students

Items

Actual count Weighted count Percent: 95%

Have you used any form of

cocaine, including powder, crack

or freebase

Yes

860

16130

1112

24922

4.3

95.7

Family give me love and support

Yes

21699

1292

23367

1412

94.3

5.7

High School: Univariate Odds Ratio. The odds of using any form of cocaine,

including powder, crack, or freebase among participants who have family that give them

love and support is 0.115 (CI: 0.095, 0.140) times that of the odds among participants

without family who give them love and support. In other words, family support is

associated with 88% lower odds of using any form of cocaine, including powder, crack,

or freebase. Since the 95% confidence interval for the odds excludes one (1), the odds are

statistically significant. Table 12 presents the odds ratio.

High School: Multivariable or Adjusted Odds Ratio. Controlling for adults over

21 around participants who have used substance in the past for HS, the odds of family

love and support among participants who indicated the used of any form of cocaine,

including powder, crack or freebase is 0.115 times the odds among participants who

indicate no used of any form of cocaine, including powder, crack or freebase. This

suggests that the use of drugs is associated with 88.5% lower odds of family love and

support. Table 12 presents the adjusted odds ratio.

Table 12

Univariate and Multivariable Odds Ratio for Assessing the Association Between Substance

Use and Social Support Among HS Students

Items Univariate odds ratio ( 95% Multivariable odds ratio

CI) (95% CI)

Dependent: Have you

used any form of cocaine,

including powder, crack

or freebase

Family give me love

and support Yes vs. No

0.115 (0.095, 0.140)

0.115 (0.083, 0.160)

Note. The multivariable logistic regression adjusted for the following question: About

how many adults over 21 have you known personally who in the past year have used

marijuana, crack, cocaine, or other drugs?

Middle school: Frequency. Of the 18,693 weighted participants of MS,

approximately one percent (1.4%) of the participants indicate during the past 30 days

took a prescription drug (such as Oxycodone, Percocet, Vicodin, Codeine, Adderall,

Ritalin, or Xanax) without a doctor’s prescriptions, while 98.6% indicated during the past

30 days not a prescription drug (such as Oxycodone, Percocet, Vicodin, Codeine,

Adderall, Ritalin, or Xanax) without a doctor’s prescriptions. About ninety-seven percent

(97.4%) of the participants indicated family gives them love and support, whereas 2.6%

have indicated not having family give them love and support. Table 13 presents the

frequency of the participants.

Table 13

Frequency of Prevalence of Substance use and Social Support Among MS Student

Items

Actual count Weighted count Percent: 95%

Family give me love and support

Yes

12458

435

6273

169

97.4

2.6

During the past 30 days, did you

take a prescription drug (such as

Oxycodone, Percocet, Vicodin,

Codeine, Adderall, Ritalin, or

Xanax) without a doctor’s

prescriptions

Yes

275

17212

259

18434

1.4

98.6

Middle School: Univariate Odds Ratio. The odds of taking a prescription drug

(such as Oxycodone, Percocet, Vicodin, Codeine, Adderall, Ritalin, or Xanax) without a

doctor’s prescription during the past 30 days is 0.124 (CI: 0.084, 0.182) times that of the

odds among participants who have family that give them love and support. In other

words, family love and support are associated with 88% lower odds of taking a

prescription drug (such as Oxycodone, Percocet, Vicodin, Codeine, Adderall, Ritalin, or

Xanax) without a doctor’s prescription during the past 30 days. Since the 95% confidence

interval for the odds excludes one (1), the odds are statistically significant. Table 14

presents the odds ratio.

Middle School: Multivariable or Adjusted Odds Ratio. Controlling for adults

over 21 around participants who have used substance in the past for MS, the odds of

using prescription drugs among participants who indicated their family gives them love

and support is 0.122 times the odds among participants who indicate their family did not

give them love and support. This suggests that family love and support are associated

with 87.8% lower odds of using prescription drugs. Table 14 presents the adjusted odds

ratio.

Table 14

Univariate and Multivariable Odds Ratio for Assessing the Association Between Substance

Use and Social Support Among MS Students

Items

Univariate odds ratio ( 95%

CI)

Multivariable odds ratio

(95% CI)

Dependent: During the

past 30 days, did you take

a prescription drug (such

as Oxycodone, Percocet,

Vicodin, Codeine,

Adderall, Ritalin, or

Xanax) without a doctor’s

prescriptions

Family give me love

and support Yes vs. No

0.124 (0.084, 0.182)

0.122 (0.061, 0.244)

Note. The multivariable logistic regression adjusted for the following question: About

how many adults over 21 have you known personally who in the past year have used

marijuana, crack, cocaine, or other drugs?

Summary

This chapter presents the result of the study. Using the MIYHS survey, the study

assessed whether an association exists between mental health problems, substance use,

and social support. Chapter 4 presents the participant’s demographics, data collection,

study results, and summary.

The study focuses on MS and HS students in all Maine’s public schools and quasi-

public schools and private schools that have about 60% of public funds. The majority of

the participants were white. The original survey used stratified random sampling to select

participants; this ensures that the participants are selected proportionate to the number of

participants in Maine. Many questions in the MIYHS survey were taken from other

national surveys, including the YRBS. The questionnaires were tested and retested and

were found to be valid and reliable. To ensure that the design features are maintained

during analysis, all estimates were weighted to reflect the distribution of the variables in

the middle and high school student population. Multivariable logistic regression was used

to assess whether an association exists between dependent and independent variables, and

this regression analysis enables adjustment for possible confounders. The result shows

that there is a statistically significant association between mental health and substance

use, mental health and social support, substance use and social support.

Chapter 5 will provide the study overview, including the findings, limitations,

recommendations, implications, and the potential for future research.

Chapter 5: Discussion, Conclusions, and Recommendations

Using secondary data from the MIYHS 2017, the purpose of my study was to

determine (a) the relationship between mental health problems and substance use among

youth who are MS or HS students, (b) the relationship between mental health problems

and social support in youth who are MS or HS students, and (c) the relationship between

substance use and social support in youth who are MS or HS students.

For my dissertation, I used a cross-sectional, retrospective design using data from

the MIYHS. I adopted the same definitions used by the MIYHS. The definitions and

supporting questions are listed below. Mental health problem was measured as a sad

feeling or suicidal ideation for the past 2 weeks. Social support was measured by love and

support from family as well as communication with family and friends. These are

captured by the following two questions: (a) I have a family that gives me love and

support (b); I have parents who are good at talking with me about things. In this survey,

substance use was measured using the following question:

• During your life, how many times have you used any form of cocaine,

including powder, crack, or freebase.

Chapter 5 provides an overview of the study, which includes the interpretation of

findings, limitations of the study, recommendations, implications, and conclusions. I will

also describe how the results of the study may help with positive social change. I used the

MIYHS secondary data to obtain the result for the study. These data were collected by the

Maine CDC using the MIYHS. Mental health problems are associated with substance use

and social support.

Interpretation of Findings

The study findings are presented in the context of the literature review on mental

health association with substance use and social support. The findings indicate that

mental health has been a major public health concern that needed further examination

among youths. In the current study, I focused on the association between mental health

problems, social support, and substance use among MS and HS students. The analysis

was conducted using SPSS version 25 for assessing the following questions:

Mental Health

•Item 14: (Have you ever seriously thought about killing yourself)?

•Item 18: (During the past 12 months, did you ever feel so sad or hopeless

almost every day for two weeks or more in a row that you stopped doing some

usual activities)

•Item 20: (During the past 12 months, did you ever seriously consider

attempting suicide).

Substance Abuse

•Item 32: (Have you ever used an electronic vapor product)?

•Item 45: (About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs)?

•Item 63: (During the past 30 days, how many times did you take a prescription

drug (such as Oxycodone, Percocet, Vicodin, Codeine, Adderall, Ritalin, or

Xanax) without a doctor’s prescription?

•Item 69: (During your life, how many times have you used any form of

cocaine, including powder, crack or freebase),

•Item 85: (About how many adults over 21 have you known personally who in

the past year have used marijuana, crack, cocaine, or other drugs)?

Social Support

•Item 107: (I have support from adults other than my parents).

•Item 108: (I have a family that gives me love and support).

Theoretical Framework

Chapter 2 described mental health problems and the association with substance use

and social support among young adults. Two conceptual frameworks were used in this

study: The relation regulation theory and self-medication theory. The self-medication

theory helps to better understand factors that make individuals to use substance. This

theory is based on the idea that an individual’s use of drug is due to an untreated

underlying problem, which could be an underlying mental health problems in the context of

this study. The self-medication is in the context of regulation, where an individual’s sense

or understanding of regulation matches self-medication. The second conceptual framework

is relational regulation theory, which helps to better understand social support with mental

health problems. The idea behind RRT is that the association between social support and

mental health problems originates from people regulating their emotions through shared

activity and conversation.

The results from this study suggest that the use of any form of cocaine, including

powder, crack, or freebase, is associated with higher odds of seriously considering

attempting suicide in the past 12 months among youths. The results of this study also

suggest that using an electronic vapor product is associated with higher odds of serious

thought about killing oneself. These findings are consistent with previous literature that

showed that substance use may be associated with a mental health problem such as

suicide (Mason et al., 2014). The association between mental health and substance use

can be explained using the self-medication theory. The self-medication theory proposes

that the use of drug among youths can lead to mental health problems (Duncan, 1974a).

Substance use is a problem worldwide, and it is a risk factor for many health problems.

Considering the association between substance use and mental health, there is a need to

implement programs and policies that discourage substance use among youths.

The current study finds that support from adults other than parents is associated

with lower odds of feeling sad or hopeless almost every day for 2 weeks or more in the

past 12 months among youths. These findings are consistent with previous literature that

stated that providing social support in the family and the community helps decrease

distress and promote hope (Holden et al., 2015). However, the current study focused on

youth, whereas previous research focused on adulthood. The association between mental

health and social support can be explained using RRT. This theory states that individuals

improve their mental health through the diversity of relationships (John & Louise, 2013),

and social support is an example of diversity of relationships.

In this current study, the results show that family support is associated with lower

odds of substance use including any form of cocaine, including powder, crack, or

freebase. The result of this study also suggests that family love and support are associated

with lower odds of taking a prescription drug (e.g., Oxycodone, Percocet, Vicodin,

Codeine, Adderall, Ritalin, or Xanax) without a doctor’s prescription. These findings are

consistent with previous literature that suggests that college students with no positive

social networks are at higher risk for using a substance, including alcohol, tobacco, and

marijuana. Support from spouse and family may be protective to prevent substance use

(Mason et al., 2014; Studer et al., 2017). This study extends the body of knowledge about

association between social support and substance use because the current study focused

on youth, whereas previous studies focused on college students and older adults. The

RRT and the self-medication theory may be used to explain the association between

social support and substance use.

Limitations of the Study

This is a descriptive, correlational, cross-sectional study that measures the

association between mental health with substance use and social support in youths

between the ages of MS and HS years. The identification of association does not mean

there is causation between the variables (Seema, 2018). This is a limitation because there

might be other possible explanations for any observed association. For example, an

unknown genetic factor may be associated with mental health, and the same genetic

factor may associate with substance use. Therefore, the observed association might be

due to an unknown confounding factor. To address the limitation of confounding factors,

the study controlled for covariates that may be associated with independent and

dependent variables, for example, social, economic status, and family composition. There

may be bias due to self-reporting because the participants may forget, deceive, or may not

understand the questions, which may lead to underreporting or overreporting (Maine

CDC, 2017). Nonetheless, studies have shown that surveys completed by young adults

could provide valid and reliable information that is needed for research (Maine CDC,

2017).

Recommendations

Because the study used a cross-sectional, retrospective, correlational design,

longitudinal and cohort studies are recommended for future study with the MS and HS

population. The study focused on the analysis of the MS and HS students’ data from the

MIYHS. The study was focused on MS and HS students in Maine. A future

crosssectional study with other states is recommended to compare the study results. The

literature review showed that mental health problems associated with social support and

substance use was reported in different countries and states, such as American Indians

from the Eastern Tribe (Ni et al., 2017), Oru camp near Ago-Iwoye in Ogun state (Amosu

et al., 2016).

Based on the study results, mental health problems are associated with substance

use and social support among MS and HS students. It is recommended that early social

support may help prevent mental health problems in adulthood (Holden et al., 2015). It is

also recommended that practitioners develop focused age-specific interventions for

students at risk for mental health conditions secondary to drug use and to plan prevention

programs for school-aged youths.

Implications

The study has the potential to bring about positive social change by using the

result of the analysis to help practitioners develop focused age-specific interventions for

students at risk of mental health conditions secondary to drug use and to plan prevention

programs for school-aged youths. Mental health problems and substance use are

associated with social support. Therefore, individuals who have compromised social

support such as family and friends may experience the consequence of substance use that

may affect their well-being (Staton et al., 2007). The information from this study will add

to the existing body of knowledge on how the prevention of substance use and the

promotion of social support could improve mental health outcomes and promote positive

social change.

By identifying mental health problems early in a young person’s life, intervention

may be implemented early, thus helping the youth with mental health problems become

more productive members of the society (Costello, 2016), and thereby promoting positive

social change. According to the University of Minnesota (2016), social support can

enhance the quality of an individual’s life by providing a buffer in times of crisis. The

information from this study could help in revising and establishing programs and using

strategies that reflect community needs and monitoring outcomes. The results could be

used to identify the strengths and challenges facing young people in society (MIYHS,

2017). Early intervention will help the youth live a more productive life and avert future

negative effects of mental health problems, including the cost of hospitalization and

youth inflicting harm on themselves (Costello, 2016).

Conclusions

Previous studies have shown an association between mental health problems,

social support, and substance use. Individuals who have compromised social support such

as family and friends may experience the consequence of substance use that may affect

their well-being. However, this study showed that identification of mental health

problems early in a young person’s life, intervention may be implemented early to help

youth become more productive in society. Findings from the study showed that lack of

social support might lead to substance use and affect the well-being of MS and HS youth.

Future research in a different state is needed in validating the finding of the study and

more research on mental health problems associated with substance use and social

support should be conducted in the future.

Association Between Mental Health problems, Substance Use, and Social Support in Youth

Students also viewed