The stress-mental health relationship

The stress-mental health relationship: social support and physical activity as moderators
in adults with Intellectual Disabilities.
THESIS
Presented in Partial Fulfillment of the Requirements for the Degree Master of Arts in the
Graduate School of The Ohio State University
By
Haleigh Morgan Scott
Graduate Program in Psychology
The Ohio State University
2012
Master's Examination Committee:
Susan M. Havercamp, Advisor
Marc J. Tassé
Betsey A. Benson
Copyrighted by
Haleigh Morgan Scott
2012
Abstract
Previous research has demonstrated a relationship between stress and mental health in
both the general population and individuals with intellectual disability (ID). In the
general population social support and physical activity have both been shown to have an
ameliorating effect upon this relationship, but little research has addressed this topic in
adults with ID or examined the gender differences that may be present in social support.
This study examines the effects of social support, gender, and physical activity on the
stress-mental health relationship in adults with ID. A nationally representative sample
from a preexisting dataset, the National Core Indicators was used to examine these
factors. Mental health was conceptualized as both mental illness and behavior problems,
as both play an important role in the overall health and wellness of individuals with ID.
Hierarchical regression was used to examine the predictive power of social support,
gender and physical activity on both mental illness and severity of behavior problems.
Stress was a significant predictor of both mental illness and severity of behavior
problems, with each additional stressor increasing the odds of poor mental health by
20%. This relationship held, even after controlling for level of ID, gender and place of
residence. Though a lack of social support was a strong predictor of having a mental
illness, individuals who lacked social support were twice as likely to have a mental
illness, it was not a predictor of behavior problems nor did gender affect social support.
ii
Physical activity was a predictor of behavior problems but not a predictor of mental
illness. The results of this study suggest that stress is an important variable that should be
considered a part of the assessment of both mental illness and behavior problems. This
study also suggests that while both mental illness and behavior problems are similarly
impacted by stress, social support may play a different role in each of these factors.
Implications for treatment and assessment of mental health concerns are discussed and
directions for future research suggested.
iii
Acknowledgments
I would like to thank my committee, especially my advisor, for their guidance and
helpful suggestions throughout this project.
Also thanks to HSRI and NASDDDS for the use of the NCI data and their
patience and aid in this project.
Lastly I would like to thank my parents for their ongoing support throughout my
life and Andrew Fetzer for his unwavering love and patience.
iv
Vita
2006................................................................Cloverleaf High School
2009................................................................B.A. University of Toledo
2010-2011 ......................................................University Fellowship
2011 to present ..............................................Graduate Teaching Associate, Department
of Psychology, The Ohio State University
Publications
Scott, H. M., & Havercamp, S. M. (in press). Measurement Error. In: Volkmar, Fred R.
(Ed). Encyclopedia of Autism Spectrum Disorders. Springer, New York.
Scott, H. M., & Havercamp, S. M. (in press). Mental Retardation. In: Volkmar, Fred R.
(Ed). Encyclopedia of Autism Spectrum Disorders. Springer, New York.
Fields of Study
Major Field: Psychology
v
Table of Contents
Abstract…………………………………………………………………………………ii
Acknowledgements……………………………………………………………………..iv
Vita……………………………………………………………………………………...v
List of Tables…………………………………………………………………………....vii
Chapter 1: Introduction………………………………………………………………...1
Chapter 2: Methods…………………………………………………………………….15
Chapter 3: Results……………………………………………………………………....21
Chapter 4: Discussion…………………………………………………………………...28
References………………………………………………………………………………36
Appendix A: Social Support……………………………………………………………43
Appendix B: Physical Activity…………………………………………………………44
Appendix C: Stressful Life Events……………………………………………………...45
Appendix D: Behavior Problems………………………………………………………..48
Appendix E: Mental Illness……………………………………………………………..49
Appendix F: Tables……………………………………………………………………...50
vi
List of Tables
Table 1: Demographics of Full and Self-Report Sample…..……………………………51
Table 2: Prescription Rationale and Presence or Absence of Mental Health Concerns…52
Table 3: Pearson Correlations between Medication Type and Mental Health…….…….53
Table 4: Correlations between Mental Illness and Types of Behavior Problems….…….54
Table 5: Logistic Regression: Stress and Mental Illness- Full Sample………….………55
Table 6: Ordinal Regression: Stress and Behavior Problems- Full Sample..……………56
Table 7: Logistic Regression: Mental Illness-Self-Report Sample………………………57
Table 8: Ordinal Regression: Behavior Problems- Self-Report Sample………………...58
vii
Chapter 1: Introduction
What is the relationship between stress and mental health in adults with
intellectual disabilities (ID)? Can social support or physical activity moderate this
relationship? Do these factors affect men and women differently? This study explores
these questions in a large, nationally representative sample of adults with ID. Although a
relationship has been shown in the ID population between stressful life events and both
problem behaviors and mental illness, the role of protective factors has not been
examined. The psychological benefits of social support and exercise have been welldocumented in the general population; however, little is known about the benefits for
people with ID. The role that gender plays in the stress-mental health relationship has
received some attention in the general population but scant research in the ID population.
This study sheds light on mental health problems in adults with ID, specifically exploring
risk and protective factors. It is expected that higher stress will predict high rates of
mental health problems. Social support and physical activity are expected to attenuate
the strength of that relationship in both men and women. Women are predicted to have
higher overall levels of social support and experience a more powerful protective effect
of social support compared to men.
Mental Illness
People with ID are more vulnerable to mental health problems than the general
population, this co-occurrence of ID and mental health is known as “dual diagnosis”.
1
One of the first prevalence studies, the Isle of Wight study, found that individuals
with ID were at a higher risk than the general population to develop psychiatric disorders
(Rutter, Tizard, Yule, Graham & Whitmore, 1976). Prevalence rates of mood disorders
in the ID population have varied by study. A 4.6% prevalence rate was found in Spain
(Salvador-Carullla, Rodriguez-Blazquez, de Molina, Perez-Marin, & Velazquez, 2000),
10.9%-50% in the United States (Charlot, Doucette & Mezzacappa, 1997; Cooper,
Smiley. Morrison, Williamson, & Allen, 2007) and in England one study found rates of
14% (Taylor, Hatton, Dixon, & Douglas, 2004). This discrepancy in prevalence may be
due to variation in study design and the use of small samples of convenience. Overall,
adults with ID are at least as likely as the general population to be diagnosed with mental
illness; and yet mental health disorders may well be under diagnosed in this population.
Rojahn & Tassé (1996) found that individuals who have a mild or moderate degree of ID
are more likely to be diagnosed with a mood disorder than individuals with severe or
profound levels of ID. This corresponds to the fact that mental illness is difficult to
diagnosis in individuals with severe and profound ID. Many factors may complicate the
diagnostic process such as severity of disability, inability to refer oneself for help, lack of
clear communication abilities, different symptom presentation, diagnostic
overshadowing, and the inexperience most professionals have with assessing psychiatric
disorders in patients with ID.
Receiving a diagnosis of a mental illness does not guarantee appropriate
treatment. Barriers to appropriate mental health treatment are arguably just as
detrimental to people with ID as barriers to receiving a diagnosis. Though psychotherapy
2
approaches in the general population have been well-researched and shown to be
efficacious, people with ID rarely receive these types of treatments and research
examining efficacy is scarce (Butz, Bowling, & Bliss, 2000). Mental illness in the ID
population tends to be treated exclusively with psychotropic medications. The heavy use
of psychotropic medication has been well-documented to the extent that people with ID
are considered to be one of the most heavily medicated populations (Reiss & Aman,
1997). Psychotropic medication is often prescribed not only to treat a diagnosed mental
illness, but also to treat behavior problems (Ghosh, Arulrajan, & Baldwin, 2010; Santosh
& Baird, 1999). In addition, efficacy and side effect profiles have not been extensively
researched for this population (Matson & Mahan, 2010). For these reasons it is important
to understand the risk and protective factors in the development and maintenance of
mental health problems in this population.
Behavior Problems
Behavior problems can include a wide variety of behaviors that are problematic to
others. This can include aggression towards self or others, destruction of property, and
stereotyped behavior. The estimated rate of behavior problems in the ID population
varies greatly by study with prevalence estimates ranging from 10-45%. Variations in
prevalence rates may be due to differences in how behavior problems are defined, how
they are measured, and who is providing the ratings. A large population based sample
found that rates of problem behaviors as rated by psychiatrists was 22.5%, while ratings
based upon a standardized measure completed through interview with a parent/caregiver
were 18.7% (Jones et al., 2008). A lower prevalence was found, 10-15%, when behavior
3
problems were defined as any behavior that required controlling measures or created a
management problem as reported by staff members (Emerson et al, 2001). Grey et al.
(2010) assessed behavior problems in adults with mild to moderate levels of ID using a
rating scale specific to behavior problems, and found that the rates of behavior problems
were 45%, though 37% of these problems were considered “less severe”.
Problem behaviors often occur in clusters together, with an individual often
exhibiting several related behaviors (Tustin, Kent, Bond, & Haskell 1991). Rates of
problem behaviors tend to be higher in individuals with lower IQ and lower adaptive
behavior (Emerson et al. 2001), and among individuals living in residential settings
(Jones et al., 2008). Kiernan and Alborz (1996) followed 34 young adults with problem
behaviors over a 5 year period and found that rates of problem behaviors stayed fairly
consistent, with 70-96% of problem behaviors being maintained at similar frequency
throughout the 5-year study period. Given the difficult nature of problem behaviors and
their tendency to be pervasive, research on problem behaviors and their correlates is
needed.
Many questions have been raised regarding the cause of behavior problems. One
study showed that the rates of behaviors problems tended to be higher in individuals who
also have epilepsy, however, this study is difficult to interpret because it did not control
for functional level (Matthews, Weston, Baxter, Felce, & Kerr, 2008). Many studies
attribute problem behaviors to a lack of communication ability and suggest that it may
itself be a form of communication. Functional behavior analysis is often used to
determine the “function” or communicative intent of behavior problems such as escaping
4
a task, or attracting a caregiver’s attention, or obtaining a desired object such as food.
When some form of communication, often through functional communication training, is
taught behavior problems decrease (Hunt, 1998; Durand & Merges, 2001). This suggests
that behavior problems may stem from a lack of efficient communication ability. A
large-scale population study found correlations between problem behaviors and
communication ability across all ages, suggesting that behavior problems may stem from
an inability to communicate health issues such as physical illness or other forms of
distress (Bott, Farmer, & Rohde, 1997; McClintock, Hall, & Oliver, 2003). This is not a
complete explanation however, because many individuals with problem behavior are able
to communicate, and some behavior problems occur in the absence of any clear
antecedent or function (McClintock, Hall, & Oliver, 2003).
Behavior Problems and Mental Illness
The relationship between problem behaviors and mental illness is complicated
because the distinction between the two is often unclear. Mental illness is notoriously
difficult to diagnose in individuals with ID, especially in those with severe to profound
ID (Rojahn & Tassé, 1996). While individuals with a mild to moderate degree of ID may
have adequate communication skills and be able to report mental illness symptoms,
individuals with more severe ID may lack the capacity to inform others of their internal
states. For example, one symptom of depression is “feelings of guilt”. Individuals with
ID may not have the language or insight to report this symptom, which contributes to the
tendency of mental health symptoms to be under-identified in ID. The symptoms
presented tend to be attributed to the ID without due consideration of co-occurring
5
physical or emotional issues, known as diagnostic overshadowing. Problem behaviors
are not a characteristic of having ID yet health care professionals that lack experience in
this area may assume that these behaviors are merely symptomatic of the disability and
fail to consider or explore other explanations. Taking into consideration the many
difficulties associated with assessing mental illness in the ID population, it has been
proposed that behavior problems may be a behavioral equivalent of psychiatric symptoms
(Fletcher, Loschen, Stavrakaki, & First, 2007).
Though several case studies have reported behavior problems being related to an
underlying mental illness, research on the topic has been mixed. One study showed a
decrease in both depressive symptoms and maladaptive behaviors following treatment for
depression (Ross & Oliver 2003). Grey et al. (2010) found that rates of severe problem
behaviors were very similar to rates of mental illness and that individuals who displayed
severe behavior problems were eight times more likely to meet cutoff scores for having a
mental illness. However, mental illness can exist without behavior problems and
behavior problems without mental illness. Ross and Oliver (2003) showed that a
significant portion of individuals diagnosed with depression do not exhibit any problem
behaviors, similarly that levels of aggressive behaviors do not differ significantly in
depressed and nondepressed groups. Current clinical guidelines suggest that when
problem behaviors have no clear function or show sudden changes in frequency or
presentation, physical and mental health explanations should be explored. Given this link
between problem behaviors and mental illness, it is important to understand the factors
that contribute to the development and maintenance of problem behaviors in ID.
6
Mental Health and Stress
Stressful life events have been consistently linked to a number of mental health
problems such as depression (Kessler, 1997), psychosis (Bebbington, Bowen, & Ramana,
1997), substance abuse (Kilpatrick, et al., 1997), as well as physical health issues (Craig
& Brown, 1984; Niaura, & Goldstein, 1992) in the general population.
At one time, it was widely assumed that individuals with ID could not experience
mental health problems because they lacked the intellectual capacity. Similarly, people
with ID were assumed to lack the capacity to feel or be adversely affected by stress.
Today research has disproved both these assumptions. In fact, individuals with ID may
be at a greater risk for experiencing stress than their counterparts (Hatton & Emerson,
2004) with fewer resources available to help cope with that stress (Lunsky & Benson,
2001). Individuals with ID seem to have similar reactions to stress as the general
population (Hatton & Emerson, 2004). This does not just apply to individuals with
higher levels of functioning. Chaney (1996) showed that individuals with profound ID
showed a physiological stress response to changes in their environment. People with an
intellectual disability report levels of stress similar to that of people with other types of
disabilities and to the general population (Bramston & Mioche, 2001; Bramston, Fogarty,
& Cummins, 1999). Stress is an important determent of physical and mental health in the
general population, so too it is important for individuals with ID.
A variety of terms have been used in the literature to describe life events that can
contribute stress including life events, stressful life events, or negative life events. For
this study, the term “stressful life events” will be used to describe events including
7
bereavement, moving, unemployment, problems at work or home, and a variety of other
situations. Previous research with 47 depressed adults with mild ID, found that these
adults experienced more stress, had frequent stressful social interactions, and used fewer
active coping skills compared to matched samples (Hartley & MacLean, 2009).
Individuals with ID who have experienced a recent loss rated higher on scales of mental
health symptoms and problems. Similarly, current mental health symptoms were
associated with higher levels of stressful life events (Hubert-Williams & Hastings, 2008).
Martorell et al. (2009) looked at the presence or absence of an ICD-10 diagnosis in
relation to both traumatic life experiences and stressful life events in a sample of 177
adults with mild or moderate ID. Traumatic life experiences were events such as
physical or sexual assault, natural disaster, or a life threatening illness. Stressful life
events consisted of events such as bereavement, the end of a relationship, loss of a job, or
moving residences. Both traumatic and stressful life events were significant predictors of
an ICD-10 diagnosis. A similar study with a larger sample size (n >1,000), found that the
presence of one or more stressful life events in the previous 12 months increased the odds
ratio for affective disorders in a population of adults with ID (Hastings, Haton, Taylor, &
Maddison, 2004). A similar study found that this relationship held, even after controlling
for all demographic variables (Owen et al., 2004). In a study of 151 adults with mild to
moderate ID, the number of life events experienced in the previous six months was a
predictor of current depression (McGillivray & McCabe, 2007). In conclusion, research
has consistently shown that stress and stressful life events have a similar effect upon the
mental health of people with ID as it has in the general population.
8
Behavior Problems and Stress
The impact of stress on problem behaviors is very similar to the relationship
between stress and mental health. Research in this area is sparse, but generally
consistent. One prospective study found that problem behaviors were predicted by
frequency counts of stressful life events (Esbensen & Benson, 2006). Several correlation
studies found that individuals who scored higher on measures of behavior problems were
more likely to have experienced stressful events in the past 6 months (Monaghan & Soni,
1992), or 12 months (Guaziuddin, 1988; Owens et al. 2004). These studies have
established evidence of a strong association between stressful life events and problem
behaviors. Further investigation into this relationship could be helpful in predicting and
preventing problem behavior.
Social Support
Social support has a significant impact on physical and mental health status and
quality of life. Social support is defined as the extent to which an individual has a
network of friends and family who can provide an outlet for frustrations and fears and
give assistance and encouragement in times of difficulty. Social support has been
extensively researched and found to play a protective role across many different
situations. Social support is negatively correlated with depression, even after stressful
life events (Aneshensel & Stone, 1982; Bell, Leroy, & Stephenson, 1982). Social
support has also been shown to aid in recovery from severe mental illness (Hendryx,
Green, & Perrin, 2009). In ID, higher levels of social support are positively correlated
9
with higher quality of life and negatively correlated with depression (Lunsky & Benson,
2001; Reiss & Benson, 1985; Meins, 1993).
Two theories regarding the relationship between social support and positive
outcomes in the face of stress have been proposed, the main effects model and the
buffering model. Research has found support for both models. The buffering model
proposes that having a high level of social support provides a barrier or buffer between
stress and poor mental health outcomes. The main effects model proposes that social
support has a general positive impact upon mental health. Evidence for the buffering
model is seen in situations where social support is defined as the availability of resources
which are needed in times of stress. For example, having a best friend or close family
members available in times of need will protect from the effects of a stressor. Evidence
for the main effects model is seen when social support is defined as how integrated an
individual is into a social network, for example how active an individual is in their
community or workplace. Though community involvement may have a positive effect
overall, it is less likely to provide help in times of need (Cohen & Wills, 1985). Research
suggests that both these models are important and continue to find evidence consistent
with each model dependent upon study design and type of social support measured
(Kornblith et al., 2001; George, Blazer, Hughes & Fowler, 1989). In this study, we will
examine the potential buffering effects of social support, as we are conceptualizing social
support as the availability of resources in a person’s life. Our study will explain the
availability of supports and what factors may serve to prevent negative outcomes when
stressful life events occur.
10
Social Support and Gender
Though social support positively affects both men and women, there are
important gender differences. In general, women tend to have higher levels of social
support, larger social networks and more emotional support than men (Belle, 1987;
Kendler, Myers & Prescott, 2005). However, when low levels of social support are
present, the negative effects are seen more strongly in women. A multinational study
found that women who had recently experienced stressful life events combined with a
lack of social support were more vulnerable to depression than men with similar levels of
social support and stress (Dalgard et al., 2006). A study which followed over 1,000
opposite sex twin pairs over a year-long period found similar results; low social support
in women predicted depression more strongly than in men (Kendler, Myers, & Prescott,
2005). Similarly, stressful life events combined with low social support was predictive of
problem behaviors in adolescent girls but not boys (Windle, 1992). Very little research
has looked how gender differences may affect social support in individuals with ID. The
only study which examined psychosocial risk factors for depression in the ID population
found that low social support from one’s family was a risk factor for women but not for
men (Lunsky, 2003).
Physical Activity
The physical benefits of exercise are well-documented and have received
considerable attention as a public health intervention. A lack of physical activity
contributes to obesity and a large number of adverse medical conditions such as diabetes,
coronary heart disease, stroke, and premature death (Warburton, Nicol, & Berdin, 2006).
11
An exciting area of research is how physical activity may benefit not only physical health
but mental health. Apart from the reported “runner’s high” or an “endorphin rush” after a
workout, research has shown that physical activity lowers the risk of clinical depression,
and lowers levels of anxiety (Fox, 1999). Several large meta-analyses have shown that
physical activity may have a positive effect on both depression and anxiety (Craft &
Lander, 1998; Strohle, 2009). A pilot study found some evidence for the benefit of
physical activity in the treatment of bipolar disorder (Ng, Dodd, & Berk, 2007).
Physical activity also serves as a buffer between life stress and negative physical
and mental outcomes. In the elderly, physical activity slowed the rate of functional
decline and buffered the effect of widowhood on functional decline (Unger, Johnson &
Marks, 1997). In studies of adolescences (Brown & Siegel, 1998; Norris, Carroll &
Cochran, 1991) and adults (Carmack, Boudreaux, Amaral-Melendez, Brantley, & de
Moor, 1999) the psychological and physical effects of stress were decreased by high
levels of physical activity. Overall, research has shown that physical activity is an
important part of maintaining physical and mental health in the general population.
The same health benefits and risks were found in samples of adults with ID
(Johnson, 2009). Few studies have examined the psychological benefit of physical
activity in this population but initial studies have been promising. Participation in an
exercise program increased feeling of social well-being and increased positive feelings
regarding physical appearance in adults with ID (Carmeli, Vaknin, Morad, & Merrick,
2005). A similar study found that participation in an exercise program led to higher life
satisfaction and lower rates of depression in adults with Down syndrome (Heller, Hsieh,
12
& Rimmer, 2004). Unfortunately, despite the benefits of physical activity on mental and
physical health, research has consistently shown that individuals with ID have a lower
level of physical fitness and lower levels of physical activity when compared to the
general population (Rimmer, Heller, Wang, &Valerio, 2004; Frey, 2004; Havercamp,
Scandlin, & Roth, 2004). Further investigation into the impact of physical activity on
mental health is important as it may help adults with ID become more resilient to stress
and its adverse effects.
Hypotheses
Hypothesis 1: Stressful life events will account for a significant portion of the variance
associated with a mental health diagnosis. Specifically, as stressful life events increase,
the likelihood of a mental health diagnosis will increase.
Hypothesis 2: Stressful life events will account for a significant portion of the variance
in severity of behavior problems. Specifically, as stressful life events increase, the
severity of behavior problems will increase.
Hypothesis 3: The diagnosis of a mental illness will have a strong correlation with
severity of behavior problems.
Hypothesis 4: Social support will act as a moderator to the stress- mental health
relationship for both mental illness and problem behaviors. As levels of social support
increases, the strength of the relationship between stress and mental health will decrease.
Hypothesis 5: Physical activity will act as a moderator to the stress- mental health
relationship for both mental illness and problem behaviors. For individuals who engage
13
in physical activity, the relationship between stress and mental health will be weaker than
for those who do not engage in physical activity.
Hypothesis 6: Women will report more social support then men.
Hypothesis 7: Social support will be a stronger moderator of the stress-mental health
relationship for women than it will be for men.
14
Chapter 2: Methods
Measures
The National Core Indicators (NCI) is a quality management protocol for the
developmental disabilities service delivery system. The NCI was created by the Human
Services Research Institute and the National Association of State Directors of
Developmental Disabilities Service to assess several key outcomes such as consumer
satisfaction, family satisfaction, cost, health, and safety.
The NCI consists of three parts: first, case managers provide demographic
information and medical record information from case files (Part I). Secondly adults
with disabilities are asked questions in a face-to-face interview where only answers from
that individual are accepted (Part II); and lastly, if the individual with DD is unable or
unwilling to continue, the final section of the survey can be completed by an interview
with a family member or someone who “knows him/her well” (Part III). This approach
of using records, self-report, and a third-party responder is supported by research done by
Lunsky, Emery & Benson (2002) which showed that the most accurate reports of health
and health care utilization were obtained when multiple sources were consulted.
Interviewers receive standardized training and states follow identical protocols to ensure
reliability across states.
Participating states collect data yearly on a random sample of at least 400 adults
with DD so that benchmarks for standards of care can be established and goals for
15
improvement set. States can measure their progress against states of similar size or
composition. The large sample size and random selection of participants required by the
NCI protocol puts it in a unique position to be used in research and policy change.
Social Support
Our social support measure consists of five NCI self-report questions. These five
questions are presented in Appendix 1. Alternate phrasing of these questions is provided
when needed to facilitate understanding, making them an effective way to obtain
perceived levels of support. For example, one question asks “Do you have a best friend
or someone you are really close to?” The question can also be phrased “Is there someone
you can talk to about personal things?” This allows the individual to report anyone that
they feel fills this role regardless of their relationship. This is a strength of the NCI as
research has shown that individuals with ID may receive social support from caregivers
or supportive environments instead of a peer or family (Lunsky & Neely, 2002). It is
important that this information is provided by the person with a disability because
caregivers cannot know the internal emotional states or perceived social support of the
adult with ID. Research has shown that while caregivers may be able to accurately report
supportive family members or caregiver’s relationships, they are less certain regarding
friendships and partners (Lunsky & Benson, 1999). Our social support scale is
represented by a value between 0 (full support) and 1 (no support). These scores are a
representation of the proportion of questions which an individual indicated no support.
For example, an individual who indicated support in 4 out of the 5 questions would have
16
a social support score of .2. Missing scores were replaced with the mean scores of the
answered social support questions.
Physical Activity
Physical activity is addressed in the NCI case report section in a question asking
about regular moderate physical activity. The specific question can be found in
Appendix 2.
Stressful Life Events
The NCI provides a measure of stressful life events consisting of 13 questions that
came from all three response sections of the NCI survey. Items were selected according
to previous research on stressful life events in the ID field and published life event scales
such as the Lifestress Inventory (Fogarty, Bramston, & Cummins, 1997). Our thirteen
items and the section of the NCI in which they can be found are listed in Appendix 3.
Items were scored on a yes/no basis. Scores range from 0, meaning that person had not
experienced any of the stressful life events listed, to 13, meaning that person had
experienced every stressful life event possible from our list.
Problem Behaviors and Mental Illness
Measures of mental illness and problem behaviors are also included in the NCI.
Mental illness is addressed by a checklist of other diagnosed disabilities (1=diagnosis, 0=
no diagnosis). The NCI addresses problem behaviors in both type and severity. Behavior
problem scores range from 0, person shows no problem behaviors, to 6, person has
behaviors severe enough to require constant intervention for all three types. Both of
17
these questions come from the case report section of the NCI. The specific questions are
listed in Appendix 4 and 5, respectively.
Participants
Participants for this study were drawn from the National Core Indicators (NCI)
database which includes 11,599 adults with developmental disabilities for survey year
2009-2010. Per NCI protocol, participants were randomly selected from a state service
registry and invited to participate as part of DD services quality control. This sample
includes individuals from 25 states- Alabama, Arizona, Arkansas, California, District of
Columbia, Florida, Georgia, Hawaii, Illinois, Kentucky, Louisiana, Maine,
Massachusetts, Missouri, New Hampshire, New Jersey, New Mexico, New York, North
Carolina, Ohio, Oklahoma, Pennsylvania, South Dakota, Texas, and Washington. The
term “developmental disability” is an umbrella term that is very heterogeneous. A
sample from the DD population could include individuals without ID diagnosed with an
Autism Spectrum Disorder, or Cerebral Palsy and these individuals may be impacted
differently by stress or have different social support and physical activity needs than an
ID population. For this reason, as well as to better define our sample and be consistent
with existing research, only participants with an ID diagnosis, as indicated by a question
on the survey, were used in this study. After selecting only those subjects diagnosed with
ID, our total sample consists of 10,627 participants.
As discussed previously, one feature of the NCI database is its use of multiple
responders. Questions which come from the self-report (Part II) section of the survey are
only able to be answered by the participant in question. If the participant chooses not to
18
answer or if they are unable to answer, the question is left blank. Though the self-report
section is a unique strength of the NCI, it creates a subsample with a portion of missing
values. The social support variable, as it contains only items from the self-report section
of the survey, is highly susceptible to missing data. For this reason a second dataset was
created. This sample contains participants who answered a minimum of three out of the
five social support questions. If a participant answered at least three questions, any blank
values on the social support measure were replaced with the mean of the other responses.
This second dataset yielded a sample size of 6,604 and will be referred to as the selfreport sample. Table 1 compares the demographic characteristics of the full and selfreport samples. As would be expected, the two groups differed significantly on level of
ID. The self-report sample is more heavily weighted towards those with higher levels of
functioning, as these are the individuals more likely to respond in a verbal interview
situation. Similarly, individuals living in institutions are underrepresented in the selfreport sample. This is likely linked to level of functioning as those with higher needs
tend to be in institutional settings. Though unavoidable, the difference between groups
has implications for the application of results to the ID population as a whole.
Study Strengths
The NCI database provides a large and representative sample and has a careful
protocol that was developed specifically for adults with ID. This protocol provides
information from individuals across a wide range of living situations, levels of ID, and
age. Due to the difficult nature of recruiting and obtaining information from adults with
ID, many studies in the field use small samples of convenience. One strength of the NCI
19
is that the participants were randomly selected from a registry of all individuals receiving
DD services. Truly random samples are very rare in the field of ID. Using the NCI
provides a large and representative sample on interesting and relevant data, which would
otherwise be nearly impossible to obtain.
20
Chapter 3: Results
Mental health problems have been conceptualized by two components: mental
illness and problem behaviors. To test the effects of stress on mental health problems,
two separate analyses were conducted: one with mental illness as the dependent variable
and one with behavior problems as the dependent variable. Logistic regression was used
to examine the relationship of stress on mental illness. Age, type of residence, and level
of ID were entered in the first step of the regression equation and stress in the second
step, to determine the unique contribution of stress in explaining variance in mental
illness after controlling for demographic information and level of ID. In a third step,
physical activity and social support were tested as moderators of the stress-mental illness
relationship by the creation of interaction terms consisting of the independent variable
and moderator (stress X social support), as modeled by Baron and Kenny (1986). The
self-report sample was used to test the impact of social support. Gender differences in
the moderating power of social support were examined by adding a three-way interaction
term (stress X social support X gender) as a fourth step. Again the self-report sample
was used in this analysis.
Ordinal logistic regression was used to examine the effects of stress on behavior
problems. Age, type of residence, and level of ID were entered as the first step of the
regression model and stress entered as the second. The third and fourth steps of the model
consisted of adding interaction terms to test our moderators, as outlined previously.
21
Mental Illness
In our sample, 36.6% of adults with ID were reported to have a co-occurring
mental illness. 37% of men had a diagnosed mental illness compared to 36.4% of
women. Adults living with family members had the lowest rate of reported mental illness
(20.5%), followed by those living semi/fully independently (40.1%), those living in
institutions (40.3%), and adults living in group homes had the highest reported rates of
mental illness (44.3%). Consistent with previous research, those with mild ID had the
highest reported rates of mental illness (44.4%), followed by moderate (37.3%), severe
(32.9%), and profound (26.8%). 34% of those with autism reported comorbid mental
illness, as did 14.9% of adults with Down syndrome.
Of those reported as having a mental illness, 86.2% were taking at least one
psychotropic medication and 13.3% were taking four or more different medications.
28.4% of adults not reported as having a mental illness were on at least one form of
psychotropic medication, with 2.4% taking four or more. Table 2 gives a breakdown of
individuals on each type of medication.
Diagnosis of a mental illness and number of medications reported were
significantly correlated, r =.527, p <.001. Table 3 breaks down the different types of
medication use reported and the correlation to having a mental illness diagnosis. Use of
medication for mood disorders was the most highly correlated to diagnosis of a mental
illness, r =.495, p <.001.
22
Behavior Problems
In the full sample, 45% of adults had at least some degree of behavior problems
with slightly more men (48.9%) than women (43.2%). Adults living with family
members had the lowest rate of reported behavior problems (28.6%), followed by those
living semi/fully independently (36%), those living in institutions (56.3%), and finally
adults living in group homes had the highest reported rates of behavior problems (57%).
Adults with mild ID had the lowest reported rates of behavior problems (40.3%),
followed by moderate (46.4%), profound (49.1%), while severe ID had the highest rates
of behavior problems (56.1%). Individuals with autism had a particularly high rate of
behavior problems (66.7%), while adults with Down syndrome had the lowest reported
rates (34.8%).
Psychotropic medication use is not limited to adults with psychiatric diagnosis.
Among adults with behavior problems, 73.5% were taking at least one form of
medication and 12% were taking four or more. Of adults with no behavior problems
28.6% were taking at least one psychotropic medication and 1.6% of adults were taking
four or more. Table 2 gives a breakdown of individuals on each type of medication by
behavior problems verses no behavior problems.
Number of medication taken and severity of behavior problems were significantly
correlated, r =.500, p <.001. Table 3 presents the correlations among types of medication
to severity of behavior problems. As would be expected, use of medication for behavior
problems was the most highly correlated to severity of behavior problems, r =.509, p
<.001.
23
Mental Illness and Behavior Problems
Of adults with a mental illness diagnosis, 66.1% had co-occurring behavior
problems compared to 34.4% of adults with no mental illness diagnosis with behavior
problems. Diagnosis of a mental illness and severity of behavior problems were
significantly correlated, r =.325, p = <.001. As presented in Table 4, correlations among
types of behavior problems varied with disruptive behavior correlating highly with
mental illness, r =.304, p <.001, and with destructive behaviors, r =.631, p <.001.
Severity of behavior problems was a significant predictor of having a mental illness after
controlling for gender, level of ID, and place of residence, Wald χ2(1) =677.17, p <.001.
Mental Illness and Stress
The relationship of stress and mental illness was examined on the full sample of
adults with ID using a hierarchical logistic regression (Table 5). The overall model fit of
the first step was significant, χ2 (1) =751.48, p <.001, indicating several variables were
predictors of mental illness. Level of ID was significant, with mild ID having the highest
risk of diagnosis followed by moderate, severe and profound. Place of residence was
also significant; adults living with family had the lowest risk followed by adults living
semi/fully indepedently, living in a group home, and living in an institution. Stress was
entered as a predictor of mental illness in Step 2. Stress was a significant predictor of a
mental illness after controlling for gender, ID level, and residence, and adding stress to
the regression equation significantly improved the predictive power of the model, χ2 (1)
=45.82, p <.001.
24
Behavior Problems and Stress
An ordinal regression analysis was used to examine the relationship of stress and
severity of behavior problems (Table 6). The overall model fit of the first step was
significant, χ2 (1) =839.32, p <.001. Gender was a significant predictor of behavior
problems, as was level of ID. Adults with mild ID had the lowest risk of more severe
behavior problems and those with severe ID had the highest. Place of residence was a
significant predictor of behavior problems; adults living with family had the lowest risk
followed by adults living semi/fully independently, living in a group home, and living in
an institution. Stress was entered as a predictor of severity of behavior problems in Step
2. Stress was a significant predictor of more severe behavior problems even after
controlling for gender, ID level, and residence. The addition of stress to the regression
equation significantly improved the predictive power of the model, χ2 (1) = 51.38,
p
<.001.
Social Support
All social support analyses were conducted on the self-report sample only. A
one-way ANOVA was used to test for differences in social support scores by gender,
diagnosis of a mental illness, and presence of behavior problems. Contrary to our
hypothesis, no differences were seen between men and women on our social support
measure, F(1, 6,187) =3.11, p =.078. There were significant difference between adults
with and without a mental illness diagnosis, F(1, 6,187) =51.32, p <.001 and those with
and without behavior problems, F(1, 6,187) =30.04, p<.001. Adults with either behavior
25
problems or a mental illness diagnosis reported lower levels of social support than their
counterparts.
The Stress-Mental Illness Relationship and Social Support
Logistic regression was used to examine the effects of social support on the
stress-mental illness relationship in the smaller self-report sample (Table 7). The first
two steps of the equation were identical to the previous analysis performed on the full
sample and yielded similar results. Step 1 introduced demographic variables and was a
significant predictor of mental illness, χ2(7)=466.79, p <.001. Step 2, introduced stress,
adding significant predictive power to the model, χ2(1)=38.18, p <.001.
Social support
was added in Step 3 of the equation as both a main effect variable and as an interaction
term with stress. The step was significant χ2 (2)=28.38, p <.001, and while social
support was a significant predictor of mental illness, Wald χ2(1)=19.45, p <.001, it did
not interact significantly with stress.
The Stress-Behavior Problems Relationship and Social Support
The role of social support in the stress-behavior problems relationship was
examined using ordinal logistic regression on the self-report sample (Table 8). The first
two steps of the equation were identical to the previous analysis performed on the full
sample and yielded similar results. Step 1 contained control variables and was a
significant predictor of behavior problems, χ2(7)=637.83, p <.001 . In step 2, stress
added significant predictive power to the model, χ2(1)=80.32, p <.001.
Social support
was added in Step 3 as both a main effect variable and as an interaction term with stress.
26
Though the step as a whole was marginally significant χ2(2)=7.74, p =.021, neither social
support nor the interaction term was a significant predictor of behavior problems alone.
Social Support and Gender
An interaction term (gender X social support) was added to each regression
equation to assess whether social support was moderated by gender. The interaction term
was not significant in the mental illness model nor was it significant in the behavior
problems model.
Physical Activity
Approximately half of our sample reported engaging in moderate levels of
physical activity (52.5%). In order to assess the impact of physical activity on mental
health, physical activity was added in the third step in the hierarchical regressions on
mental illness and behavior problems, which consisted of demographic variables (Step 1)
and stress (Step 2). Physical activity was not a predictor of mental illness, nor did
physical activity moderate the stress-mental illness relationship (Table 5). Physical
activity was a significant predictor of behavior problems, Wald χ2(1)=10.86, p=.001.
Physical activity did not significantly moderate the stress-behavior problem relationship,
Wald χ2(1)=2.77, p=.096 (Table 6).
27
Chapter 4: Discussion
The stress-mental health relationship was examined in a large nationally
representative sample of adults with ID using an existing NCI dataset. Stress was a
significant predictor of poor mental health, even after controlling for demographic
variables, including gender, level of ID and place of residence. Social support and
physical activity also played roles in explaining variance in mental health after
controlling for both demographic variables and stress. Results have implications for the
prevention, assessment and treatment of mental health problems in individuals with ID.
Mental Health
Mental health was conceptualized as two variables in this study, mental illness
and behavior problems. We feel that this is a more complete way to examine the mental
health of this population, as both mental illness and problem behaviors can be extremely
detrimental to the quality of personal relationships and overall quality of life. In this
sample, the prevalence of mental illness was approximately 36%, this aligns fairly well
with previous research and adds to the body of research demonstrating that mental illness
in the ID population exceeds that of the general population. The overall prevalence of
behavior problems was 45%, a finding consistent with the current literature.
The relationship between problem behaviors and mental illness has been
inconsistent across studies. In our sample individuals with mental illness were more
likely to display behavior problems (66.1% vs. 34.4%). Severity of behavior problems
28
was a significant predictor of having a mental illness diagnosis, where a one-step change
in behavior problem severity led to a 60% increase in the odds of being diagnosed with
mental illness. Behavior problems were significantly correlated with mental illness and
accounted for approximately 9% of the variance associated with having mental illness.
This provides evidence for an association between behavior problems and mental illness,
specifically as behavior problem severity increases. However, the results of this study
show that behavior problems often occur in the absence of a mental illness diagnosis and
vice versa. There are multiple explanations for the association between these variables.
It has been suggested that some behavior problems manifest as behavioral equivalents for
psychiatric symptoms, which may be driving some of the association between these
variables. It is also possible that some of the mental illness diagnoses in our sample were
inappropriately made based solely on the existence of behavioral difficulties. Future
research should continue to examine these constructs in unison and investigate
fluctuations over time and in relation to treatment. In addition, research should seek to
differentiate between behavior problems which may be associated with a mental illness
and those which may be related to other factors in an individual life, such as an
undiagnosed physical problem.
Both diagnosis of mental illness and the presence of behavior problems were
significantly correlated with psychotropic medication use. Medication use had a very
similar relationship with both behavior problems and mental illness with diagnosis of a
mental illness accounting for 25.7% of the variance in psychotropic medication use and
severity of behavior problems accounting for 25%. This is problematic for several
reasons. Many of the medications being used to treat behavior problems lack research on
29
their efficacy for these purposes, and people with ID may respond differently to
medications than the general population, again an area which has seen minimal research.
This is particularly troubling since medication is the most common form of treatment for
people with ID, while other approaches, such as psychotherapy and behavioral
interventions may be underutilized in favor of medication. The use of medication in this
population is a topic that needs a great deal of additional research on both the efficacy of
frequently prescribed medications and on alternative options, such as psychotherapy.
Prescription guidelines for ID should also be developed and implemented to ensure that
there is a standard of best practice for medication use in this population (Reiss & Aman,
1998).
Stress-Mental Health Relationship
A strong relationship between stress and negative mental health outcomes was
found. For every additional stressor reported, the odds of having a mental illness
diagnosis increased by 20%. Similarly, each additional stressor increased the odds of
having more severe behavior problems by 19%. This relationship was similar in the selfreport sample, with each stressor increasing the odds of mental illness by 20% and the
severity of behavior problems by 28%. Although significant differences between the
two samples are evident; the samples conformed to similar patterns on the variables of
interest. This lends support for our practice of generalizing the results of the social
support analyses to the larger sample.
It is of interest to note that the effect of stress on behavior problems is greater in
the self-report sample, which is largely comprised of higher-functioning individuals,
while the effect on mental illness did not differ between samples. It becomes
30
increasingly difficult to diagnose mental illness as severity of ID increases, and this may
be an important difference between our two samples. A higher rate of mental illness was
found in the self-report sample (40.1% compared to 36.6%) and a lower rate of behavior
problems (40.8% compared to 45%). The greater influence of stress on behavior
problems in the self-report sample may suggest that stress is being experienced or
reported differently by individuals with higher levels of functioning. It is also possible
that behavior problems may function differently at different levels of functioning.
Individuals with more severe ID may be more likely to display behavior problems related
to an undiagnosed physical or mental health concern, in which case we would not expect
stress to have as much of an impact. Those with less severe ID are more likely to be able
to adequately request and be treated for physical or mental health concerns and so may be
more likely to display behavior problems as a function of stress.
Though mental illness and behavior problems seem to respond in a similar fashion
to stress, they are not one and the same construct. The results of this study suggest that
stress is an important variable that should be considered a part of the assessment of both
mental illness and behavior problems. Stress should be considered a risk factor mental
illness and preventative measures should be taken when an individual is experiencing a
stressful time in his/her life. Current treatment approaches for behavior problems are
often centered on behavioral interventions and medication. Adding stress to a behavior
model could help identify the cause of behavior problems or help identify individuals
who are at risk for developing behavior problems. Identification of contributing factors
would allow for preventative steps to be taken with high-risk individuals. This also has
31
implications for treatment of behavior problems. Behavioral or cognitive-behavioral
approaches may be well suited to teaching skills for coping with stress.
Social Support and Physical Activity
We predicted that social support would act as a moderator of the stress-mental
health relationship; however, our findings did not support this. In the behavior problem
analysis, social support neither predicted behavior problems directly nor did it interact
with stress to predict behavior problems. Neither did social support moderate the
relationship between stress and mental illness. Social support did, however, have a strong
main effect on mental illness. Individuals reporting no social support were twice as
likely to have a diagnosis of mental illness compared to adults with adequate social
support (defined as having support in each of our 5 domains). Though this study did not
find evidence for the buffering model of social support it did provide strong evidence for
the main effect model. Despite social support not behaving in quite the manner expected,
it clearly is playing a significant role in the mental health of people with ID.
Social support accounted for a significant amount of the variance in mental
illness. In fact, the impact of social support was more powerful than stress in explaining
mental illness. This has serious implications for prevention and treatment of mental
illness. Social support measures may help to identify those at risk for developing a
mental illness and should be considered when looking at factors that contribute to the
maintenance of mental illness. Social support considerations should weigh heavily on
treatment plan decisions. Further research, especially prospective studies, are needed to
further the knowledge on how social support may be related to mental illness in this
population.
32
Though our analysis suggested that both mental illness and behavior problems are
negatively impacted by stress, the results of this study suggest that social support impacts
behavior problems and mental illness differently. While social support was a strong
predictor of mental illness, it did not seem to play a role in behavior problems. Further
research is needed to replicate and explain this finding.
It is important to note that our social support analyses were conducted on a subset
of the full sample and therefore may not be representative of the population of adults with
ID. Since our social support measure relied on self-report, these findings may be most
representative of adults with less severe ID. Further research is needed to explore social
support in adults with severe ID. One potential area of future research is to explore ways
to measure social support in this largely underrepresented group.
While physical activity has been shown to have a positive effect on mental illness
in the general population, no effect was found in this sample. Though possible that
physical activity does not provide the same benefit for people with ID, it seems more
likely that this result can be attributed to other factors. Most research that established the
buffering effects of physical activity on mental illness has focused on mood disorders,
and the effect of physical activity on other disorders, for example schizophrenia, has not
been shown. We were not able to test our hypotheses on mental illness sub-samples
because mental illness was a dichotomous variable.
While physical activity did not explain variance in mental illness it was a
significant predictor of behavior problems. Contrary to expectations, the presence of
physical activity was associated with an increase in behavior problems. Report of regular
physical activity was associated with a 19% increase in severe behavior problems. This
33
was an unexpected finding, and since it does not seem likely that engaging in physical
activity increases severity of behavior problems, other explanations must be considered.
One likely explanation is that a third variable, age, may be largely driving the effect.
Severity of behavior problems has been shown to decrease with advancing age (Emerson
et al., 2001) and engagement in physical activity also tends to decrease as individuals
age. Physical limitations may also be affecting this relationship. Further analysis of the
data showed that age and physical activity were significantly negatively correlated, as
was severity of behavior problems and age.
Limitations
Since our study is not a longitudinal design, we are not able to draw causal
conclusions from our results. Though stress was a strong predictor of poor mental health
outcomes, we can’t know if it was a cause, a consequence or a correlate. It is likely that a
bidirectional relationship exists with stress both contributing to and being created by
mental health problems. Studies that examine this relationship with a prospective design
are needed to further explore this area. The mental illness variable used in this study was
not ideal. Presence of absence of a mental illness was determined by a review of case
records and was not confirmed by a diagnostic professional. The use of an existing
dataset is also limiting in that the survey is pre-determined and was not written with our
specific questions in mind. For example, we would ideally prefer additional information
on specific mental illness diagnosis, physical activity and nature of social support;
however, that information was not included in the NCI survey so it was not available.
Additionally it is important to note that the NCI data is collected from a population of
34
adults receiving DD services. This study cannot generalize to the many adults with ID in
the community who are not receiving services.
Strengths of Study
Despite its’ limitations, this study uniquely contributed to the field of dual
diagnosis in several ways. First, this study looks at several variables of great interest to
the dual diagnosis field in a large sample of adults in the US receiving DD services. This
is extremely rare, and most studies of stress and mental illness exclusively focus on
individuals in the mild/moderate range of ID. This study was able to demonstrate that
large scale survey data can be adapted for use by researchers as a compliment to smaller
prospective studies to in order to better examine topics of interest. Secondly, the study
provided an initial look at physical activity as a variable associated with mental health in
the ID population. Lastly, the study showed the importance of social support for mental
health and provided the base for future studies to further explore social support and stress
as factors affecting mental health in people with ID.
35
References
Aneshensel, C. S., & Stone, J. D. (1982). A test of the buffering model of social support.
Archives of General Psychiatry, 39(12), 1392-1396.
Baron, B. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51(6), 1173-1182.
Bebbington, P., Bowen J., & Ramana R. (1997). Life events and psychotic disorders. In:
Clinical disorders and stressful life events (ed. T. W. Miller) International
Universities Press, Madison, Connecticut.
Belle , D. (1987). Gender differences in the social moderators of stress. In R. C. Barnett
L. Biener G. K. Baruch (Eds.), Gender and stress (pp. 257-277). New York: Free
Press.
Bell, R. A., Leroy, J. B., & Stephenson J. J. (1982). Evaluating the mediating effects of
social support upon life events and depressive symptoms. Journal of Community
Psychology 10(4), 325-340.
Bott, C., Farmer. R., & Rohde J. (1997). Behavior problems associated with lack of
speech in people with learning disabilities. Journal of Intellectual Disabilities
Research, 41(1) 3-7.
Bramston, P., Fogarty, G., & Cummins, R. A. (1999). The nature of stressors reported by
people with an intellectual disability. Journal of Applied Research in Intellectual
Disabilities, 12(1), 1-10.
Bramston, P., & Mioche, C. (2001). Disability and stress: a study in perspectives.
Journal of Intellectual and Developmental Disabilities, 26(3), 233-242.
Brown, J. D., & Siegel, J. M. (1988). Exercise as a buffer of life stress: A prospective
study of adolescent health. Health Psychology, 7(4), 341-353.
Butz, M. R., Bowling, J. B., & Bliss, A. C. (2000). Clinical experience and effective
clinical practice - Psychotherapy with the mentally retarded: A review of the
literature and the implications. Professional Psychology, research and practice,
31 (1), 42-47.
36
Carmack, C. L., Boudreaux, E., Amaral-Melendez, M., Brantley, P. J., & de Moor, C.
(1999). Aerobic fitness and leisure physical activity as moderators of the stressillness relation. Annals of Behavioral Medicine, 21(3), 251-257.
Carmeli, E., Zinger-Vaknin, T., Morad, M., &Merrick, J., (2005). Can physical training
have an effect on well-being in adults with mild intellectual disability.
Mechanisms of Aging and Development, 126 (2), 299-304.
Chaney R. H. (1996). Psychological stress in people with profound mental retardation.
Journal of Intellectual Disability Research, 40(4), 305-310.
Charlot, L. R., Doucette, A. C., & Mezzacappa, E. (1997). Affective symptoms of
institutionalized adults with mental retardation. American Journal on Mental
Retardation, 101, 445-458.
Cohen, S. & Wills, T. A. (1985). Stress, social support and the buffering hypothesis.
Psychological Bulletin 98(2). 310-357.
Cooper, S.A., Smiley, E., Morrison, J., Williamson, A., & Allan, L. (2007). Mental illhealth in adults with intellectual disability: prevalence and associated factors. The
Brtish Journal of Psychiatry, 190, 27-35.
Craig T. K. J. & Brown G. W. (1984). Goal frustration and life events in the etiology of
painful gastrointestinal disorder. Journal of Psychosomatic Research, 28, 411421.
Craft, L. L. & Lander D. M., (1998). The effect of exercise on major depression and
depression resulting from mental illness: A meta-analysis. Journal of Sport and
Exercise Psychology, 20 (4), 339-357.
Dalgard, O. S., Dowrick, C., Lehtinen, V., Vazquez-Barquero, J. L., Casey, P.,
Wilkinson, G., Ayuso-Mateos, J. L., Page, H., Dunn, G. & The ODIN Group
(2006). Negative life events, social support and gender difference in depression:
A multinational community survey with data from the ODIN study. Social
Psychiatry and Psychiatric Epidemiology, 4 (6), 444-451.
Durand M. V. & Merges E., (2001). Functional communication training: A contemporary
behavior analytic intervention for problem behaviors. Focus on Autism and other
Developmental Disabilities, 16(2), 110-119.
Esbensen, A. J. & Benson, B. A. (2006). A prospective analysis of life events, problem
behaviors, and depression in adults with intellectual disability. Journal of
Intellectual Disability Research. 50(4) 248-258.
Emerson E., Kiernan C., Alborz A., Reeves D., Mason H., Swarbrick R., Mason L. &
Hatton C. (2001). The prevalence of challenging behaviors: a total population
study. Research in Developmental Disabilities, 22, 77-93.
37
Fletcher, R., Loschen, E., Stavrakaki, C., & First, M. (Eds). (2007) Diagnostic ManualIntellectual Disability (DM-ID): A Textbook of Diagnosis of Mental Disorders in
Persons with Intellectual Disability. Kingston, NY: NADD Press.
Fogarty, G. J., Bramston, P., , & Cummins, R. A. (1997). Validation of the Lifestress
Inventory for people with a mild intellectual disability. Research in
Developmental Disabilities, 18(6), 435-456.
Fox, K. N. (1999). The influence of physical activity on mental wellbeing. Public Health
Nutrition, 2(3a), 411-418.
Frey, G. C. (2004). Comparison of physical activity levels between adults with and
without mental retardation. Journal of Physical Activity & Health, 1(3), 235-245.
George, L. K., Blazer, D. G., Hughes, D. C. & Fowler, N. (1989). Social support and the
outcome of major depression. British Journal of Psychiatry, 154, 478-485.
Ghosh, S., Arulrajan, A. E., & Baldwin, D. (September 01, 2010). Unlicensed
applications of licensed psychotropic drugs in an intellectual disability clinical
service: Retrospective case-note study. Journal of Intellectual Disabilities, 14 (3)
237-243.
Grey, I., Pollard, J., McClean, B., MacAuley, N., & Hastings, R. (2010). Prevalence of
psychiatric diagnoses and challenging behaviors in a community-based population
of adults with intellectual disability. Journal of Mental Health Research in
Intellectual Disabilities, 3(4), 210-222.
Ghaziuddin, M. (1988). Behavioral disorder in the mentally handicapped. The role of life
events. British Journal of Psychiatry, 152, 683-686.
Hartley S. L. & MacLean W. E. (2009). Depression in adults with mild intellectual
disability: role of stress, attributions, and coping. American Association on
Intellectual and Developmental Disabilities, 114 (3), 147-160.
Hastings R. P., Hatton C., Taylor J. L., & Maddison C. (2004). Life events and
psychiatric symptoms in adults with intellectual disabilities. Journal of
Intellectual Disability Research, 48(1) 42-46.
Hatton C. & Emerson E. (2004). The relationship between life events of psychopathology
amongst children with intellectual disabilities. Journal of Applied Research in
Intellectual Disabilities, 17, 109-117.
Havercamp, S., Scandlin, D., & Roth, M. (2004). Heath disparities among adults with
developmental disabilities, adults with other disabilities, and adults not
reporting disabilities in North Carolina. Public Heath Reports, 119, (418-426).
38
Heller, T., Hsieh, K., & Rimmer, J. (2002). Barriers and supports for exercise
participation among adults with Down Syndrome. Journal of Gerontological
Social Work, 38(1-2), 161-178.
Hendryx, M., Green, C. A. & Perrin, N. A. (2009). Social support, activities, and
recovery from severe mental illness: STARS study findings. Journal of
Behavioral Health Services and Research, 36 (3), 320-239.
Hubert-Williams L. & Hastings R. P. (2008) Life events as a risk factor for psychological
problems in individuals with intellectual disabilities: a critical review. Journal of
Intellectual Disability Research, 52, 883–95.
Hunt, S. T. (1998). Using functional communication training to alleviate problem
behaviors in young children. (Unpublished doctoral dissertation). Northern
Arizona University, Arizona.
Johnson, C.C., (2009) The benefits of physical activity for youth with developmental
disabilities: A systematic review. American Journal of Health Promotion 23 (3),
157-167.
Jones, S., Cooper, S., Smiley, E., Allan, L., Williamson, A., , & Morrison, J. (2008).
Prevalence of, and factors associated with, problem behaviors in adults with
intellectual disabilities. Journal of Nervous and Mental Disease, 196(9), 678-686.
Kendler, K. S., Myers, J. & Prescott, C. A. (2005). Sex Differences in the Relationship
Between Social Support and Risk for Major Depression: A Longitudinal Study of
Opposite-Sex Twin Pairs. American Journal of Psychiatry 162, 250-256.
Kessler, R. C., (1997). The effects of stressful life events on depression. Annual Review
of Psychology, 48, 191-214.
Kiernan, C. & Alborz A. (1996). Persistence and change in challenging and problem
behaviors of young adults with intellectual disability living in the family home.
Journal of Applied Research in Intellectual Disabilities, 9(3) 89-93.
Kilpatrick D. G., Acierno R., Resnick H. S., Saunders B. E. & Best C. L (1997). A 2year longitudinal analysis of the relationship between violent assault and
substance use in women. Journal of Counseling and Clinical Psychology, 65,
834-847.
Kornblith, A. B., Herndon II, J. E., Zuckerman, E., Viscoli, C. M., Horwitz, R. I.,
Cooper, M. R.,… Holland, J. C. (2001). Social support as a buffer to the
psychological impact of stressful life events in women with breast cancer. Cancer
91 (2), 443-454.
39
Lunsky, Y. (2003). Depressive symptoms in intellectual disability: does gender play a
role? Journal of Intellectual Disability Research 47(6), 417-427.
Lunsky, Y., & Benson, B. A. (1999). Social circles of adults with mental retardation as
viewed by their caregivers. Journal of Developmental and Physical Disabilities,
11(2), 115-129.
Lunsky, Y., & Benson B. A. (2001). Association between perceived social support and
strain, and positive and negative outcomes for adults with mild intellectual
disability. Journal of Intellectual Disability Research, 45 (2) 106-114.
Lunsky, Y., Emery, C. F., & Benson, B. A. (2002). Staff and self-reports of health
behaviors, somatic complaints, and medications among adults with mild
intellectual disability. Journal of Intellectual & Developmental Disability, 27(2),
125-135.
Lunsky, Y., & Neely, L. C. (2002). Extra-individual sources of social support as
described by adults with mild intellectual disabilities. Mental Retardation,
40(4), 269-277.
Matson, J. L., & Mahan, S. (November 01, 2010). Antipsychotic drug side effects for
persons with intellectual disability. Research in Developmental Disabilities, 31,
6, 1570-1576.
Martorell, A., Tsakanikos E., Pereda, A., Gutiérrez-Recacha, P., Bouras, N. & AyusoMateos J. L. (2009). Mental health in adults with mild and moderate intellectual
disabilities: The role of recent life events and traumatic experiences across the life
span. Journal of Nervous and Mental Disease, 197(3), 182-186
Matthews, T., Weston, N., Baxter, H., Felce, D., & Kerr, M. (2008). A general practicebased prevalence study of epilepsy among adults with intellectual disabilities
and of its association with psychiatric disorder, behavior disturbance and career
stress. Journal of Intellectual Disabilities Research, 52(2) 163-173.
Meins, W. (1993). Prevalence and risk factors for depressive disorders in adults with
intellectual disability. Australia and New Zealand Journal of Developmental
Disabilities, 18 (3), 147-156.
McClintock, K., Hall, S., & Oliver, C. (2003). Risk markers associated with challenging
behaviours in people with intellectual disabilities: A meta-analytic study. Journal
of Intellectual Disability Research, 47(6), 405-416.
McGillivray J. A. & McCabe M. P. (2007) Early detection of depression and associated
risk factors in adults with mild/moderate intellectual disability. Research in
Developmental Disabilities. 28(1), 59-70.
40
Monaghan, M. T. & Soni, S. (1992). Effects of significant life events on the behavior of
mentally handicapped people in the community. British Journal of Mental
Subnormality, 38(2), 114-121.
Ng, F., Dodd, S., & Berk, M. (2007). The effects of physical activity in the acute
treatment of bipolar disorder: A pilot study. Journal of Affective Disorders,
101(1-3), 259-262.
Niaura R. & Goldstein M. G (1992). Psychological factors affecting physical condition:
cardiovascular disease literature review: II. Coronary artery disease and sudden
death and hypertension. Psychosomatics: Journal of Consultation Liaison
Psychiatry, 33, 146-155
Norris, R., Carroll, D., & Cochrane, R. (1992). The effects of physical activity and
exercise training on psychological stress and well-being in an adolescent
population. Journal of Psychosomatic Research, 36(1), 55-65.
Owens D. M., Hastings R. P., Noone S J., Chinn J., Harman K., Roberts J. & Taylor K.
(2004). Life events as correlates of problem behavior and mental health in a
residential population of adults with developmental disabilities. Research in
Developmental Disabilities, 25(4), 309-320.
Reiss, S., & Aman, M. (1997). The international consensus process on
psychopharmacology and intellectual disability. Journal of Intellectual Disability
Research, 41(6), 448-455.
Reiss, S., & Aman, M. (1998). Psychotropic medications and developmental disabilities:
the international consensus handbook. Washington, DC: The American
Association on Mental Retardation.
Reiss, S. & Benson, B. A. (1985). Psychosocial correlates of depression in mentally
retarded adults: I. Minimal social support and stigmatization. American Journal
of Mental Deficiency ,89(4), 331-337.
Rimmer, J. H., Heller, T., Wang, E., & Valerio, I. (2004). Improvements in physical
fitness in adults with Down Syndrome. American Journal on Mental Retardation,
109(2), 165-174.
Rojahn, J., & Tasse M. J., (1996). Psychopathology in Mental Retardation. Washington
D.C: American Psychological Association.
Ross, E. & Oliver, C. (2003). The assessment of mood in adults who have severe or
profound mental retardation. Clinical Psychology Review, 23(2), 225-245.
Rutter, M., Tizard, J., Yule, W., Graham, P., & Whitmore, K. (1976). Isle of Wight
studies, 1964-1974. Psychological Medicine, 6, 313-332.
41
Salvador-Carulla, L., Rodriguez-Blaazquez, C., de Molina, M. R., Perez-Marin, J., &
Velazquez, R. (2000). Hidden psychiatric morbidity in a vocational program for
people with intellectual disability. Journal of Intellectual Disability Research, 44,
147-154.
Santosh, P. J., & Baird, G. (1999). Psychopharmacotherapy in children and adults with
intellectual disability. The Lancet, 354(9174), 233-242.
Strohle, A. (2009). Physical activity, exercise, depression and anxiety disorders.
Biological Psychiatry, 116(6), 777-784.
Taylor, J. L., Hatton, C., Dixon, L. and Douglas, C. (2004), Screening for psychiatric
symptoms: PAS-ADD Checklist norms for adults with intellectual disabilities.
Journal of Intellectual Disability Research, 48, 37–41.
Tustin, R. D., Kent, P. A., Bond M. J. & Haskell, H. (1991). A classification of behavior
problems as exhibited by people with intellectual disabilities. Journal of
Intellectual and Developmental Disability, 17(3), 303-312.
Unger, J. B., Johnson, C. A., & Marks, G. (1997). Functional decline in the elderly:
Evidence for direct and stress-buffering protective effects of social interactions
and physical activity. Annals of Behavioral Medicine, 19(2), 152-160.
Warburton, D. E. R., Nicol, C. W., & Bredin, S. S. D. (2006). Prescribing exercise as
preventive therapy. Canadian Medical Association Journal, 174(7), 961-974.
Windle, M. (1992). A longitudinal study of stress buffering for adolescent problem
behaviors. Developmental Psychology, 28(3), 522-530.
42
Appendix A- Social Support Measures
Section of NCI
27) Do you have friends you like to talk to or do things with?
If s/he answers "yes," ask who the friends are and try to determine if they are
family, staff, roommates, co-workers, etc. You can use prompts such as: Can
you tell me their names? Are these friends staff or your family?
__2 Yes, has friends who are not staff or family
__1 Yes, all friends are staff or family, or cannot determine
__0 No, does not have friends *
__9 Don’t know, no response, unclear response
28) Do you have a best friend, or someone you are really close to? (Is
there someone you can talk to about personal things?)
__2 Yes
__0 No *
__9 Don’t know, no response, unclear response
9) Can you see your friends when you want to see them? (Can you make
plans with your friends when you want to?)
We are trying to determine if person gets support to see friends. Try to factor
out situations where friends are not available – this is not the issue.
__8 NOT APPLICABLE – does not have any friends *
__2 Yes, can see friends whenever s/he wants to
__1 Sometimes can’t see friends (e.g., not enough staff or transportation)
__0 No, often unable to see friends *
__9 Don’t know, no response, unclear response
31) Do you ever feel lonely? (Do you ever feel like you don’t have anyone to
talk to?)
If s/he responds “yes,” probe to determine how often s/he feels lonely.
__2 [Yes] – often feels lonely (more than half the time) *(Reverse scored)
__1 Sometimes (about half the time)
__0 [No] – not often (less than half the time)
__9 Don’t know, no response, unclear response
33) Can you see your family when you want to? (Can you pick the times
you see them? Does someone help you make plans to see them?)
If family is not available or does not wish to have contact, code as NOT
APPLICABLE. If the person has family but does not want to see them, code
as 2.
__8 NOT APPLICABLE – family not available, person does not have
family or family does not wish to have contact *
__2 Yes, sees family whenever s/he wants to, or chooses not to see family
__1 Sometimes
__0 No *
__9 Don’t know, no response, unclear response
43
Part II
Part II
Part II
Part II
Part II
Appendix B-Physical Activity
Section of NCI
BI-24. Does this person routinely engage in any moderate physical
activity? (Moderate physical activity is an activity that causes some increase
in breathing or heart rate. Examples include, but are not limited to, brisk
walking, swimming, bicycling, cleaning, and gardening.) (Check ONE)
__ 1 Yes
__ 2 No
44
Part I
Appendix C-Stressful Life Events
Section
of NCI
BI-14. Overall, how would you describe this person’s health?
__1 Excellent or very good
__2 Fairly good
__3 Poor*
__4 Don’t know
BI-30. How often does this person require medical care? (Check ONE) (Medical care
refers to care that must be performed or delegated by a nurse or physician. Do not include
medication administration.)
__ 1 Less frequently than once/month
__ 2 At least once/month, but not once a week
__ 3 At least once/week, or more frequently*
__ 4 Don’t know
BI-33. How long has this person lived in his/her current home?
__ 1 Less than 1 year*
__ 2 1-3 years
__ 3 3-5 years
__ 4 Over 5 years
__ 5 Don’t know
Do you have a job in the community?
A community job refers to paid work - either competitive or supported employment
(includes both individual and group employment, such as a work crew or enclave). It does
not include work done in facility-based settings like sheltered workshops. It also does not
include volunteer work.
PS-8 (Do you work at _____________________________________?)
__2 Yes (code Question 2 as NOT APPLICABLE )
__0 No ask Question 2
__9 Don’t know, no response, unclear response
2) If No, ask: Would you like to have a job in the community?
__8 NOT APPLICABLE – has job in the community
__2 Yes*
__1 In-between
Do you have a job in the community?
A community job refers to paid work - either competitive or supported employment (includes
both individual and group employment, such as a work crew or enclave). It does not include
work done in facility-based settings like sheltered workshops. It also does not include volunteer
work.
PS-8 (Do you work at _____________________________________?)
__2 Yes code Question 2 as NOT APPLICABLE
__0 No ask Question 2
__9 Don’t know, no response, unclear response
3) Do you like working there?
45
Part I
Part I
Part I
Part II
Part II
__8 NOT APPLICABLE – no job in the community
__2 Yes
__1 In-between
__0 No*
__9 Don’t know, no response, unclear response
7) Do you go to a day program or do some other activity during the day? This does not
include a job in the community.
PS-8 (Do you go to ___________________________________?)
__2 Yes
__0 No code Questions 8-11 as NOT APPLICABLE
__9 Don’t know, no response, unclear response
8) Do you like going there/doing this activity?
__8 NOT APPLICABLE – no day program or other activity
__2 Yes
__1 In-between
__0 No*
__9 Don’t know, no response, unclear response
13) Do you like your home or where you live? (Do you like living here?)
__2 Yes
__1 In-between
__0 No *
__9 Don’t know, no response, unclear response
18) Are they (staff) nice and polite to you? (Do they treat you with respect?)
__8 NOT APPLICABLE – no home support staff
__2 Yes
__1 Sometimes or some staff
__0 No *
__9 Don’t know, no response, unclear response
22) Are you ever afraid or scared when you are at home?
__2 [Yes] - most of the time*
__1 Sometimes
__0 [No] - rarely
__9 Don’t know, no response, unclear response
23) Are you ever afraid or scared when you are out in your neighborhood?
__2 [Yes] - most of the time*
__1 Sometimes
__0 [No] - rarely
__9 Don’t know, no response, unclear response
24) Are you ever afraid or scared at work or at your day program/other activity?
__8 NOT APPLICABLE – no work or day program/activity
__2 [Yes] - most of the time*
__1 Sometimes
__0 [No] - rarely
__9 Don’t know, no response, unclear response
25) If you feel afraid, is there someone you can go to for help?
__8 NOT APPLICABLE – does not feel afraid
__2 Yes
__1 Maybe, not sure
__0 No*
__9 Don’t know, no response, unclear response
79) Do you get the services you need?
(Other respondent – Does this person get the services and supports s/he needs?)
Respondent: ( ) 1-individual ( ) 2-family/friend ( ) 3-staff ( ) 4-other
46
Part II
Part II
Part II
Part II
Part II
Part II
Part II
Part III
__2 Yes code Question 79a as NOT APPLICABLE
__1 Sometimes, or doesn’t get enough of the services needed
__0 No *
__9 Don’t know, no response, unclear response
47
Appendix D-Behavior Problems
Section of NCI
Does person need support to manage:
BI-54. Self-injurious behavior
Refers to attempts to cause harm to one’s own body, for example, by hitting
or biting self, banging head, scratching or puncturing skin, ingesting inedible
substances, or attempting suicide.
Part I
Level of Support Needed (Check ONE)
__1 No support needed
__2 Some support needed; requires only occasional assistance or
monitoring
__3 Extensive support needed; frequent or severe enough to require regular
assistance
__9 Don’t Know
Part I
BI-55. Disruptive behavior
Refers to behavior that interferes with the activities of others, for example, by
laughing or crying without apparent reason, yelling or screaming, cursing, or
threatening violence.
Level of Support Needed (Check ONE)
__1 No support needed
__2 Some support needed; requires only occasional assistance or
monitoring
__3 Extensive support needed; frequent or severe enough to require regular
assistance
__9 Don’t Know
BI-56. Destructive behavior Refers broadly to externally-directed, defiant
behavior, for example, taking other people’s property, property destruction,
stealing, or assaults and injuries to others.
Level of Support Needed (Check ONE)
__1 No support needed
__2 Some support needed; requires only occasional assistance or
monitoring
__3 Extensive support needed; frequent or severe enough to require regular
assistance
__9 Don’t Know
48
Part I
Appendix E-Mental Illness
BI-10. What other disabilities are noted in this person’s record? (Check ALL
that apply)
__ 1 Mental Illness/Psychiatric Diagnosis (e.g. Depression)
__ 2 Autism Spectrum Disorder (e.g., Autism, Asperger Syndrome, Pervasive
Developmental Disorder)
__ 3 Cerebral Palsy
__ 4 Brain Injury
__ 5 Seizure Disorder/Neurological Problem
__ 6 Chemical Dependency
__ 7 Limited or No Vision- Legally Blind
__ 8 Hearing Loss- Severe or Profound
__ 9 Physical Disability
__ 10 Communication Disorder
__ 11 Alzheimer’s Disease or other Dementia
__ 12 Down Syndrome
__ 13 Prader-Willi Syndrome
__ 14 Other disabilities not listed
__ 15 No other disabilities other than MR/ID
__ 16 Don’t know
49
Section of NCI
Part I
Appendix F
Tables
50
Table 1. Demographics of Full and Self-Report Sample
Male
Age
18-34
35-54
>55
Race/Ethnicity
White, non-Hispanic
Black, non-Hispanic
Other, non-Hispanic
Hispanic
Severity of Intellectual Disability
Mild*
Moderate
Severe
Profound**
Other Disabilities
Autism
Down Syndrome
Alzheimer’s Disease/dementia
Place of Residence
Family
Semi/Fully Independent
Group Home
Institution*
Full Sample
(10,627)
56.6%
Self-Report Sample
(6,604)
55.6%
28.4%
47.5%
24.1%
31.3%
46.0%
22.7%
70.1%
19.1%
4.9%
5.9%
70.1%
20.3%
4.9%
4.7%
35.7 %
26.6%
15.6%
22.1%
52.7%
32.9%
10.4%
4.0%
10.5%
8.2%
1.6%
7.1%
8.3%
1.3%
29.3%
17.7%
30.1%
22.8%
35.1%
23.4%
31.7%
9.9%
*Significant at the .01 level **Significant at the <.001 level
51
Table 2. Prescription Rationale and Present or Absence of Mental Health Concerns
Mental Illness
Medication
Behavior Problems
Yes
No
Yes
No
Mood Disorder
64.6%
17.1%
52.9%
20.7%
Anxiety
42.1%
14.5%
38.6%
14.8%
Psychotic Disorders
37.6%
7.1%
31.0%
8.9%
Behavior Problems
43.0%
15.3%
48.1%
7.7%
prescribed for:
52
Table 3. Pearson Correlations between Medication Type and Mental Health
Medication for Mood
Disorder
Mental Illness
.495**
Behavior Problems
.344**
Medication for Anxiety
.327**
.294**
Medication for Psychotic
Disorder
.395**
.330**
Medication for Behavior
Problems
.317**
.509**
** Significant at the <.001 level
53
Table 4. Correlations between Mental Illness and Types of Behavior Problems
Self-Injurious
Disruptive
Destructive
Self-Injurious
1
Disruptive
.484**
1
Destructive
.469**
.631**
1
Mental Illness
.188**
.304**
.264**
** Significant at the <.001 level
54
Mental Illness
1
Table 5. Logistic Regression: Stress and Mental Illness-Full Sample
55
Variable
Gender
Male
Level of ID*
Moderate
Severe
Profound
Place of Residence**
Semi/fully Ind.
Group
Institution
Stress
Physical Activity
PA/Stress
Interaction
Significance of Step
B
-.02
-.33
-.71
-1.3
.85
1.22
1.58
Step 1
Wald
OR
.163
.163
.98
326.73
32.16
.71
94.16
.49
315.60
.26
517.36
136.15 2.38
369.62
3.4
429.64 4.86
-
-
-
p
.687
.687
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
-.32
-.66
-1.24
.82
1.2
1.57
Step 2
Wald OR
.054
.054
.99
276.15
28.94
.73
81.71
.51
267.25 .28
505.42
128.03 1.97
367.57 2.95
421.67 4.13
-
.181
-
45.52
-
B
-.011
2
χ =45.82
2
2
Significance of
<.001
χ =751.48
χ =797.3
Model
*Mild ID as comparison group **Living with family as comparison group
55
1.20
-
p
.82
.82
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
.83
1.21
1.55
Step 3
Wald
OR
.022
.022
.99
275.77
29.27
.72
82.89
.51
267.07
.28
495.75
128.47 2.28
358.23 3.35
403.26 4.73
p
.88
.88
<.001
<.001
<.001
<.001
<.001
<.001
<.001
<.001
.15
-.09
.059
17.63
2.28
1.22
<.001
.13
.27
B
-.007
-.32
-.67
-1.27
χ2=2.39
2
χ =799.69
1.16
.91
1.06
.30
<.001
Table 6. Ordinal Regression: Stress and Behavior Problems-Full Sample
Variable
Gender
Male
Level of ID*
Moderate
Severe
Profound
Place of Residence**
Semi/fully Ind.
Group
Institution
56
Stress
Physical Activity
PA/Stress
Interaction
Significance of Step
B
.15
.35
.63
.068
.47
1.21
1.28
-
Step 1
Wald
OR
13.30
13.30
.16
125.72
42.49
1.42
98.69
1.88
1.13
1.07
593.22
49.27
1.60
475.16 3.36
379.94 3.58
-
-
p
<.001
<.001
<.001
<.001
<.001
.287
<.001
<.001
<.001
<.001
.45
1.2
1.27
Step 2
Wald OR
14.88
14.88 1.17
131.28
46.65 1.45
112.41 1.97
5.58
1.16
585.59
44.08 1.57
462.77 2.97
376.75 3.56
-
.17
-
52.15
-
B
.16
.37
.68
.16
2
χ =51.38
2
2
Significance of
<.001
χ =839.32
χ =890.70
Model
*Mild ID as comparison group **Living with family as comparison group
56
1.19
-
p
<.001
<.001
<.001
<.001
<.001
.02
<.001
<.001
<.001
<.001
<.001
-
.44
1.20
1.32
Step 3
Wald
OR
12.49
12.49
1.16
133.78
47.6
1.45
118.94
2.01
9.3
1.22
598.73
42.75
1.56
459.52
3.31
393.74
3.75
p
<.001
<.001
<.001
<.001
<.001
.02
<.001
<.001
<.001
<.001
.14
.18
.08
17.2
10.86
2.77
<.001
.001
.096
B
.15
.37
.70
.20
1.15
1.19
1.08
<.001
χ2=27.77
<.001
<.001
χ2=918.47
<.001
Table 7. Logistic Regression: Mental Illness- Self-Report Sample
Step 1
Step 2
Step 3
57
Variable
B
Wald
OR
p
B
Wald
OR
p
B
Wald
OR
p
Gender
Male
1.16
1.16
-.06
1.16
1.16
.281
.281
-.07
1.42
1.42
.93
.94
.281
.281
.94
-.06
.234
.23
Level of ID*
Moderate
Severe
Profound
-.30
-.70
-1.15
83.35
21.83
46.16
48.86
<.001
<.001 -.30
<.001 -.70
<.001 -1.11
83.36
20.98
45.64
44.92
.74
.50
.33
<.001
<.001
<.001
<.001
-.30
-.72
-1.20
90.12
20.83
48.14
51.10
.74
.48
.30
<.001
<.001
<.001
<.001
Place of Residence**
Semi/fully Ind.
Group
Institution
.73
1.2
1.62
339.70
82.18 2.08
268.34 3.32
238.28 5.04
<.001
<.001
<.001
<.001
.70
1.18
1.56
339.70
75.50 2.02
255.99 3.24
219.41 4.77
<.001
<.001
<.001
<.001
.69
1.16
1.53
327.21
72.69
247.99
209.70
1.70
2.76
3.76
<.001
<.001
<.001
<.001
.74
.50
.32
Stress
-
-
-
-
.18
37.82
1.20
<.001
.17
18.86
1.19
<.001
Social Support
-
-
-
-
-
-
-
-
.77
19.45
2.17
<.001
SS/Stress interaction
-
-
-
-
-
-
-
-
-.04
.13
.96
.72
Significance of Step
2
χ =38.18
2
2
Significance of
<.001
χ =446.79
χ =484.97
Model
*Mild ID as comparison group **Living with family as comparison group
57
<.001
χ2=28.38
<.001
χ =513.35
<.001
2
<.001
Table 8. Ordinal Logistic: Regression Behavior Problems- Self-Report Sample
Step 1
Step 2
Step 3
58
Variable
B
Wald
OR
p
B
Wald
OR
p
B
Wald
OR
p
Gender
Male
.02
.16
.16
1.02
.685
.685
.03
.30
.30
1.03
.58
.58
.028
.26
.26
1.03
.61
.61
Level of ID*
Moderate
Severe
Profound
.27
.45
.26
34.91
20.02
24.50
3.58
1.31
1.57
1.30
<.001
<.001
<.001
.058
.28
.46
.33
37.71
21.42
25.62
5.76
1.33
1.58
1.39
<.001
<.001
<.001
<.001
.28
.46
.30
37.18
21.56
25.42
4.83
1.33
1.58
1.37
<.001
<.001
<.001
<.001
.49
1.27
1.81
529.10
39.29
344.70
368.56
1.64
3.54
6.15
<.001
<.001
<.001
<.001
.46
1.24
1.75
502.3
33.83
329.57
342.08
1.58
3.46
5.78
<.001
<.001
<.001
<.001
.46
1.23
1.74
492.66
33.47
325.12
344.41
1.58
3.44
5.69
<.001
<.001
<.001
<.001
Stress
-
-
-
-
.25
81.58
1.28
<.001
.21
33.44
1.23
<.001
Social Support
-
-
-
-
-
-
-
-
.21
1.76
1.23
.18
SS/Stress interaction
-
-
-
-
-
-
-
-
.12
1.31
1.12
.25
Place of Residence**
Semi/fully Ind.
Group
Institution
2
Significance of Step
Significance of Model
χ =80.32
2
χ =637.83
2
<.001
χ =718.16
*Mild ID as comparison group **Living with family as comparison group
58
<.001
χ2= 7.74
<.001
χ =725.9
.02
2
<.001