RISK FOR PSYCHOPATHOLOGY AND RATINGS OF LABORATORY
STRESS: ARE THEY RELATED?
Tasha Nicole Aper
B.A., Saint Louis University, 2002
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
This thesis for the Master of Arts
Tasha Nicole Aper
has been approved
David S. Albeck
Aper, Tasha Nicole (M.A. Psychology)
Risk for Psychopathology and Ratings of Laboratory Stress: Are They Related?
Thesis directed by Assistant Professor David S. Albeck
This study asked whether cumulative risk for psychopathology and ratings
of laboratory stress are related. Participants in the study were fifty-five 17-19 year
olds (27 males, 28 females). The sample was predominantly Caucasian and middle
class. Cumulative risk scores for psychopathology were calculated at 0-2, 3-4,6,
and 12 years with retrospective longitudinal data. Individuals participated in the
Trier Social Stress Test, after which they completed a rating of the stressor: the
Post Trier Social Stress Test (PTSST). Self-report measures of current functioning,
including the Multidimensional Anxiety Scale for Children (MASC), the
Childrens Depression Inventory (CDI), the Youth Self-Report (YSR), and the
UCLA PTSD Index were collected. Regression analyses were utilized to examine
relationships between cumulative risk, current functioning, and ratings of the social
stressor. Cumulative risk scores did not predict PTSST scores, nor did they predict
total scores for current functioning. However, current functioning predicted 25%
of the variability in PTSST scores, F(4,36) = 4.28, p = .006. Specifically, total
scores on the MASC were positively related to PTSST scores, [B = .341, p = .001].
Because the population was subclinical, relationships between subscales on
measures of current functioning and PTSST scores were examined with Pearson r
correlations. The Physical Symptoms (r = .31, p = .023) and Social Anxiety (r =
.448, p = .001) subscales of the MASC were positively to PTSST scores.
Furthermore, several YSR subscales were positively related to PTSST scores,
including the Anxious Depressed (r = .49, p = .000), Social Problems (r = .27, p =
.05), and Withdrawn subscales (r = .27, p = .05). Although limitations to the study
include retrospective data, small sample, and subclinical population, there are
implications for the findings. That is, individuals with deficits in current
functioning may perceive certain laboratory stressors as more threatening. This
may effect measurement of responses to stress, particularly when grouped
according to another unrelated variable. Researchers should control for current
functioning when comparing group responses to laboratory stress to avoid
attributing differences to an unrelated variable. In short, even subclinical levels of
anxiety may mediate ratings of stress.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
David S. Albeck
For my family and friends...
A special thanks to Kim Kelsay, M.D. and Mary Klinnert Ph.D. at National Jewish
Medical and Research Center for the use of their data and for their support in the
completion of this project.
Risk for Psychopathology..................................1
Environmental Risk Factors................................2
Examining Effects of Environmental Risk...................7
Clinical Rating Scales............................14
Indicators of Risk................................14
Indicators of Current Functioning............... 22
Cumulative Risk Scores..................24
3.1 Distribution of 0-2 Year Cumulative Risk Scores............................25
3.2 Distribution of 3-4 Year Cumulative Risk Scores............................26
3.3 Distribution of 6 Year Cumulative Risk Scores.............................26
3.4 Distribution of 12 Year Cumulative Risk Scores............................26
4.1 Distribution of Clinically Significant Scores on Self-Report Measures......29
Risk for Psychopathology
The notion of predicting psychopathology has been a popular topic of
research literature over the past three decades. It seems researchers and clinicians
alike long to unlock the etiology of the development of psychopathology.
Particularly, in the field of developmental psychopathology, the ability to predict
negative outcomes in mental health is paramount. Claiming someone is at risk,
for psychopathology, implies that this particular individual has an increased
probability of later exhibiting the disorder in question. This however, does not
necessarily mean that a causal relationship exists. The potential for identification
of individuals at risk for psychopathology is powerful because it provides an
opportunity for some form of intervention, which might create a positive change in
the trajectory of an individuals development. Such an opportunity has
immeasurable value, especially to the individuals who suffer from various forms of
emotional and behavioral problems. However, an overarching problem in the
identification of individuals at risk for psychopathology is the inseparable
relationship which exists between humans and their environment.
In Sameroff & Chandlers (1975) transactional model of developmental
psychopathology, both the quality of the environment and traits of the child work
together to explain behavior and adjustment. So, both child and environment are
active in shaping behaviors through adaptation. For instance, individual factors
within a child determine how the environment, particularly the parent, will respond
to the child. Temperamentally difficult children are two times as likely to receive
parental criticism (Rutter, Quinton, and Yule, 1977). In such a case, what are
considered to be ingrained characteristics of the child have the potential to shape
the responses of the caregiver. Thus, to some extent, a childs early social
environment is a result of individual characteristics, a mechanism in which genetic
factors can shape environments (Rutter, 1979). However, research also shows that
early maternal anxiety has been found to be linked to temperamental difficulty in
infants (Vaughn, Bradley, Joffe, Seifer, & Barglow, 1987; Van den Bergh, 1990).
The extraction of a causal relationship between these factors is difficult, because of
the interconnected nature of the individuals genetics and the environment.
There is strong evidence supporting the contribution of environmental risks
for psychopathology (Rutter, 2005). Over the past three decades, great strides have
been taken in the identification of risk factors for the development of
psychopathology in children and adolescents. Several research studies have
uncovered relationships between a large number of environmental or ecological
risk factors and an assortment of psychological outcomes, including
psychopathological outcomes. Ideally, the objective of this line of research is to
show a causal relationship (and its direction) between some risk factors) and
psychopathological risk. However, because prospective designs, extremely large
samples, and experimental control over environmental risks are not always
financially or ethically feasible, many research studies have not accomplished the
aforementioned goal. Again, it is also difficult to resolve whether a
psychopathological effect is a result of genes or environment, as they tend to be
correlated (Rutter, 2005). However, consistent findings across a multitude of
research studies point to a few environmental factors, now widely accepted in the
field as risks for psychopathology (Samerofif, Lewis, & Miller, 2000).
Environmental Risk Factors
In order to understand the environment surrounding a child, we must
examine the family. The literature is consistent in its description of risky
families. Such family environments generally include aspects of aggression,
conflict, and poor nurturing (Repetti, Taylor, & Seeman, 2002). With the ability to
identify environmental risk factors, we can recognize families and/or adolescents
with a propensity toward suboptimal development and/or psychopathology. Such a
task requires a clear understanding of which environmental factors lead to the
development of psychopathology.
Parental psychopathology has received much attention in the literature as a
risk factor for the development of psychopathology in children. This may be due,
in part, to the existence of a genetic component for certain forms of
psychopathology. However, detrimental parental behavioral patterns associated
with psychopathology also play a significant role. Most often, behavioral parenting
issues involve some form of harsh treatment or a lack of supervision. Diminished
mental health in parents might work to weaken effectiveness in performing
caretaker behaviors. Associations between maternal depression and less productive
infant-mother interactions are well documented. Offspring of women with
depression show increased risk for problems in emotional regulation, less favorable
interactions, insecure attachment, behavior problems, and delays in competency
development (Downey & Coyne, 1990; Gelfand & Teti, 1990). Carter, Garrity-
Rokous, Chazan-Cohen, Little, and Briggs-Gowan (2001) found that maternal
depression in combination with other psychopathological conditions and/or life
adversity raises the risk associated with parent-child relationships and upsets
childrens developmental progress. This group concluded that work with multi-risk
families should focus on minimizing depression and impairment associated with
psychopathological conditions, and also work to promote effective parenting. In a
large scale study of adolescents aged 14 to 17 years, researchers again found that
parental psychopathology, namely social phobia and depression, were associated
with the development of social phobia in their youth (Lieb et al., 2000). Such
findings are particularly relevant to the study at hand, as we have access to various
measures of maternal depression across each childs lifespan.
There is agreement in the field that, in order to examine developmental
trajectories at a young age, one must consider the caregiving context. This is due to
the dependent nature of young children on parents for nearly all basic needs. Also,
this level of regulation by caregivers gradually shifts to self-regulation as the child
progresses to more responsibility for his/her well-being (Sameroff & Fiese, 2000).
Problematic parent and child relationships, which may be greatly affected by
parental psychopathology, can negatively influence child development. Several
studies demonstrate the inter-connectedness of child behavior and parenting.
Variations in parenting strategies have been found to intensify negative behaviors.
Through experimental manipulation of parenting strategies, one study found a
relationship between inconsistent consequences for negative child behavior and
increased child negativity and demanding behavior (Acker & OLeary, 1996). Lieb
and colleagues (2000) also found a relationship between parenting styles
characterized by rejection and overprotection with the development of social
phobia in adolescents. Thus, it is possible that poor parent and child relationships,
which may reflect poor parenting contribute to the development of
Parental stress also has the potential to affect the environmental climate
surrounding a child. In a recent review article, Huizink, Mulder, and Buitelaar
(2004) presented evidence in support of the notion that exposure to prenatal stress
may result in a vulnerability to psychopathology. They utilized evidence from
animal studies, revealing overactivity and impaired negative feedback regulation of
the hypothalamic-pituitary-adrenal (HPA) axis in the offspring of prenatally
stressed animals. Such findings may indicate a physiological mechanism in the
development of psychopathology. Also, in a population of children with severe
emotional disturbance, a retrospective look at prenatal medical histories revealed
significant exposure of unplanned stressors by the mother (e.g. marital conflict,
emotional disturbances) (Ward, 1991). Severe distress has also been linked, again
retrospectively, to attention deficit hyperactivity disorder (Clements, 1992).
Furthermore, maternal anxiety has been found to be linked to temperamental
difficulty in infants (Van den Bergh, 1990; Vaughn et al., 1987). Such findings
clearly indicate some component of parental stress in the development of
psychopathology in children.
Marital distress has the potential to influence development of
psychopathology. Amato & Keith (1991) concluded that current research
substantiates claims that children from divorced families display moderate
vulnerability to problems in academic, social, psychological, and behavioral
adjustment. However, others suggest that pre-divorce variables more accurately
predict adjustment (Block, Block, & Gjerde, 1986; Emery, 1982). The marital
discord that precedes divorce may be responsible for disrupted attachment,
aggressive modeling, problematic disciplinary skills, or stress on the child that
negatively affects development (Emery, 1982). In a population of 13-year-olds,
Rueter, Scaramella, & Wallace (1999) found that, while parental disagreements did
not directly influence onset of psychiatric disorder, they were important moderators
in the increase of self-reported depressive symptoms in the population. Klinnert,
Mrazek, & Mrazek (1992) found a relationship between quality of prenatal marital
relationship and the number of behavior problems predicted at two years of age.
Furthermore, children who are exposed to marital conflict may often feel anger,
shame and fear. They might also see themselves as responsible for the conflict
(Grych & Fincham, 1993). Such recurring parental disagreements may be
responsible for producing the depressive symptoms that might lead to a clinical
A childs own psychological distress may negatively impact his/her
developmental trajectory. The importance of early detection of
behavioral/emotional problems in children has been recognized in recent years due
to an increasing emphasis on early detection and intervention (Carter, Briggs-
Gowan, & Davis, 2004). A number of children display psychopathology at an
early age, and problems developed early often persist into young adulthood. While
diagnosis of psychopathology is progressively easier as a child develops,
determining the mental health of an infant is more difficult Temperament can be
an early indicator of psychological distress. Graham, Rutter, & George (1973)
found temperamental characteristics to be strongly related to the development of a
psychiatric disorder. Several researchers concede that difficult temperament is
often related to later social competence and behavioral problems (Seifer, 2000).
So, even at an early age, psychological disturbance has the power to predict future
Low socioeconomic status (SES) has immense effects on the development
of a childs competence. Overarching poverty in a childs environment plays an
important role in the etiology and course of both psychological and behavioral
problems. Unfortunately, die effects of poverty may overlap with a multitude of
other environmental risks, making it difficult to parse out the effects of poverty
itself on the development of child psychopathology. Often problematic is the
isolation of distinct proximal environmental risk variables, which often tend to
reside together within such an overarching distal variable. Povertys effect on
parents can transform a single risk into four: financial strain or poverty, single
parenthood, parental depression, and poor parenting (McLoyd, 1998). However,
there is some evidence that encounters with adversity do not explain negative
outcomes independentiy of coexisting risk factors. Sameroff, Bartko, Baldwin,
Baldwin, & Siefer (1998) found that, when comparing children who are raised in
wealthy or poor families who had the same number of total risk factors, there was
no discrepancy in their competence. So, while poverty may be an important
environmental risk factor, we must consider the possibility of a number of risk
factors being represented by this one construct. Thus, it is important to measure the
other variables that might compose the risk within poverty, as we will do in the
While minority status is often considered a risk for psychopathology, one
must consider that this uniform word does not describe the heterogeneous groups of
people that compose this label. However, many minorities do have something in
common. That is, minority status in the United States can often mean a restricted
access to resources and power. Because of this systematic exclusion, minority
children may be placed in suboptimal developmental pathways from birth.
Minority status may often mean experiencing the cumulative effects of other
poverty-related factors such as poor prenatal care, environmental toxins, poor
nutrition, parental un- or underemployment, and many other social disadvantages.
This places them at risk for developmental psychopathology (Garda Coll &
Garrido 2000). While not always the case, minority status may also lead to
interpersonal racial discrimination, which may, in turn, lead to mental illness. In
the United States, discrimination has been associated with hypertension,
depression, stress, and poor health (Krieger, 2000). However, although racism may
predict mental illness in minority groups we must understand that this relationship
exists through such social factors as housing, benefits, and education, not in
individual weaknesses of a particular ethnic, cultural, or racial group (Chakraborty
& McKenzie, 2002).
Psychosocial risk can influence development in a multitude of ways. In a
recently published review model, Kaugars, Klinnert, & Bender (2004)
demonstrated that family emotional characteristics influenced asthma onset and
outcome in a pediatric population. While stressful conditions are associated with
increased risk for illness, illness and disease can also become a significant source
of stress in itself for children and their families (Compas & Boyer, 2001). Such
stress may lead to the development of psychopathology. Utilizing a prospective
longitudinal design Mrazek, Schuman, & Klinnert (1998) found that children who
developed asthma by age three were at greater risk for behavior problems than
children with later asthma onset and children who did not develop asthma. Thus,
the existence of a chronic illness may be associated with increased stress and,
potentially, increased behavioral disruptions in children and adolescents.
The course of human development is characterized by the introduction of
both risk and protective factors working in concert over an individuals lifetime.
At any given time, many such complex systems operate simultaneously during all
stages of development. Specifically, several aspects of a childs environment
interact to form an environmental/social climate. We must ask what type of
environment creates individuals most vulnerable to psychopathology. To examine
each individual risk factor encountered would be a daunting, if not impossible, task.
This problem can be resolved through a process of cumulative risk modeling. This
method for modeling the interactions of potential environmental risk factors
introduced to a child assists psychologists in measuring the complexities of
coexisting environmental risks.
Cumulative risk modeling involves measuring all systems affecting an
individual, including proximal and distal qualities of the social and physical
environment. In his hallmark study, Michael Rutter (1983) utilized this approach
for analyzing such complex systems affecting an organism simultaneously. A
Cumulative Risk Score is derived by calculating a sum for the multiple risk
categories identified for an individual. Cutoff points are determined for risk in a
given area and the number of times cutoffs are met is summed across a multitude of
potential environmental risk factors, similar to those mentioned above. One
method of comparison is to dichotomize individuals into classifications of risk
exposure (e.g., high or low risk) according to each individual cumulative risk score
(Rutter 1983,1993). Or, alternatively, a continuous Cumulative Risk Score can be
used in regression analyses.
One major advantage of such a model is its capacity to simultaneously
model several risk factors while avoiding pitfalls of major statistical and
interpretation liabilities of multiplicative interactions (Evans, 2003). Generally,
when compared to this simple additive model of counting stressful events, the
utilization of intricate methods to weight life events does not change correlations
between stress and disorder. This leads researchers to the conclusion that such
procedures are needless (Compas, Davis, Forsythe, & Wagner, 1987).
Examining Effects of Environmental Risk
As mentioned before, in order to understand environmental risk, we must
attempt to examine both proximal factors directly experienced by the individual
and distal factors that exert indirect consequences. As previously discussed, distal
environmental risk factors, such as poverty, are an overarching category for a
multitude of other simultaneously existing risks. The deleterious effects of such an
all-encompassing risk factor point to the importance of Cumulative Risk Modeling.
Several research studies have shown that it is not the existence of one particular
risk in a childs life that has substantial negative effects. It is the coexistence of
several risk factors working together that exerts the most negative influence on a
The Rochester Longitudinal Study (RLS; Sameroff, Seifer, Baldwin, &
Baldwin, 1993; Sameroff, Seifer, & Zax, 1982) was designed to examine the
impact of parental psychopathology on child behavior. Throughout the course of
this study, risk factors were outlined that, while prevalent in lower SES groups,
affect child outcomes in every social strata. The environmental conditions
identified as developmental risks were: 1) history of maternal mental illness; 2)
high maternal anxiety; 3) parental rigidity in attitudes, beliefs, and values regarding
child development; 4) lack of positive maternal interactions in infancy; 5) head of
household in unskilled occupations; 6) minimal maternal education; 7)
disadvantaged minority status; 8) single parenthood; 9) stressful life events; and 10)
large family size. The researchers found that a single variable such as parental
psychopathology did not determine a negative outcome. It was the accumulation of
several risk factors that caused vulnerability to negative consequences, particularly
mental health. After dividing participants in the RLS into groups of low and high
numbers of risk factors, the researchers also found that individuals in the low-risk
group displayed better mental health and cognitive outcomes at every age measured
(ages 1,4,13, and 18 years) (Sameroff, 1998).
Michael Rutter (1979) also argues the number of risk factors in a childs
environment at any given time determine mental health outcomes, not a single risk
factor. In his popular study, risk factors included: 1) severe marital distress; 2) low
SES; 3) large family size; 4) paternal criminality; 5) maternal psychiatric disorder;
and 6) child admission to foster care. For this sample of 10-year-old children,
psychiatric risk in families with zero or one risk was two percent, while twenty
percent of the children in families with four or more risks showed psychiatric risk.
That is, twenty percent of the individuals with four of more cumulative risk factors
developed a psychiatric disorder. This finding revealed how stresses can potentiate
each other, thus, presenting further evidence for the significance of the
accumulation of several risk factors at one time in child development of
psychopathology. These findings are of paramount importance because they
suggest that the elimination of some stresses, even if others cannot be removed, can
be of great value. Another important consideration is the fact that some, if not
many, of the individuals introduced to such environmental risks may emerge
unscathed. Rutter suggested this was due, in part, to the number of simultaneous
Also, in a large (N = 792) sample of 11 year-old children Williams,
Anderson, McGee, and Silva (1989) found that sex, maternal depression, parental
marital status, and reading difficulties differentiated between children with and
without behavioral and emotional disorder. The researchers recruited their children
from a normal population. So, while many faced a behavioral or emotional
disorder, it was not recognized or the children were not impaired enough to require
treatment. However, it could be important to identify moderate forms of
psychopathology in at-risk populations, as the potential for identification of risk
before a problem is salient to others may provide a good opportunity for
In New Zealand, researchers found that a cohort of fifteen year-olds with
multiple problem behaviors were often from disadvantaged, dysfunctional, and
disorganized homes, again showing concordance between the number of risk
factors encountered and the number of resulting behavior problems (Fergusson,
Horwood, Lynskey, 1994).
In a recent review article Rutter (2005) summarized research findings on
psychosocial experiences carrying risk. Three broad categories of risk emerged:
(1) lack of stable, harmonious, and committed relationships; (2) lack of social
cohesion; and (3) lack of interchange and play which should lead to significant
learning opportunities in the areas of cognitive skill development and social coping.
A comprehensive study examining the effects of environmental risk on the
development of psychopathology should include variables from each of these
Individuals suffering from various forms of psychopathology may be prone
to cognitive distortions. Those with symptoms of anxiety and depression are biased
to attend to negative aspects of the environment (Mathews & MacLeod, 1994).
Specifically, clinical anxiety has the potential to affect thinking, perception, and
learning. It can also cause confusion and create distortions in perception. These
distortions may be related to time and space and/or persons and meanings of
events. One particularly significant trait of anxious individuals processing is the
tendency to select particular features in the environment and ignore others in an
attempt to justify fear of a situation (Sadock & Sadock, 2003). Because they fear
humiliation or embarrassment in social settings, individuals with social phobia may
attend to the negative reactions of their audience, instead of positive features in
an effort to justify their fear of partaking in the social interaction. Furthermore, one
mode of psychological intervention, Cognitive Therapy, developed by Aaron Beck,
targets the modification of dysfunctional thinking that characterizes a diverse set of
psychological disorders (Beck, 1964). Thus, it is noteworthy to mention that
individuals with psychopathology attend to negative aspects of the environment.
Several mechanisms of environmental risk are plausible for the relationship
between risk and psychopathology. One such model involves the development of
negative or maladaptive cognitive sets. Dodge, Pettit, Bates, & Valente (1995)
found that children who were victims of abuse were more likely to display hostile
attributional biases toward videos depicting social situations that ranged from
antagonistic to benign. The same individuals were also prone to externalizing
A similar concept, cognitive appraisals of stressful situations, has been
linked to health (Lazarus & Folkman, 1984). This was tested by Chen & Matthews
(2003) through the use of ambiguous videos in a population of low SES
adolescents. They found that low SES was associated with threatening
interpretations of two ambiguous videos of social situations. Similar results were
also reported in a previous study, where low SES children suggested more
threatening qualities than high SES children in response to ambiguous stories
(Chen & Matthews, 2001).
Because low-SES adolescents are more likely to be exposed to stressful
situations and environments (Brady & Matthews, 2002), they may perceive the
world as a threatening place and be more apt to label ambiguous situations as
threatening (Chen & Matthews, 2003). While this is an important contribution, we
have previously discussed the implication of utilizing SES as a general risk factor
for maladaptive behaviors. Low SES can indicate the presence of several risks
In the current study, the accumulation of various environmental risk factors
in a predominantly middle class population was examined. By measuring
cumulative risk of environmental factors that can exist independently of low SES, it
was possible to examine the effect of cumulative risk on appraisals of stressful
situations. Also, the participants were subjected to a novel laboratory stress
situation. They were part of a stressful situation, and did not simply examine a
video and imagine themselves as the protagonist in an artificial situation. This was
a unique opportunity to examine how past and current experiences influenced
ratings of a stressful situation.
As part of the W. T. Grant Asthma Risk Study (Klinnert et al, 1992;
Klinnnert et al, 1994; Mrazek, Klinnert, Mrazek, & Macey, 1991; Mrazek, Mrazek,
& Klinnert 1995; Mrazek, et al., 1998), 150 children in the Metropolitan Denver
area who were the offspring of women with asthma, and were thus at risk for
developing asthma, were followed since the prenatal period to the age of seventeen
to nineteen. Extensive biological and psychosocial measures were taken during the
prenatal period, frequently during the first four years of life, at age 6 and at age 12.
Furthermore, between the age of 17 and 19,60 participants were recruited and
exposed to a social stress test At that time, various measures of current
psychological functioning were taken. The cumulative risk scores created for this
study were the first such scores developed for this sample of individuals.
The available data lended itself to the examination of longitudinal risk,
current functioning, and appraisals/ratings of a stressful situation. This study
specifically investigated possible differences in participants ratings of the stressful
event. This was interesting for several reasons. First, the sample was primarily
Caucasian and middle class. Thus, there was an opportunity to examine risk that
might not have been the result of poverty and/or chronic adversity. Also, because
measures of past risk and current functioning were available, it was possible to
investigate whether past risk or current functioning was more relevant to ratings of
1. Cumulative risk scores (the number of simultaneous risks) at each time
period (years 0-2, years 3-4,6 years, and 12 years) will predict Total
2. Problem scores in current functioning, as measured by self-report measures
(the MASC, CDI, UCLA PTSD index, and YSR) will predict ratings on the
Total PTSST Scores.
3. Finally, cumulative risk scores at each time period (years 0-2, years 3-4,6
years, and 12 years) will predict problems in current functioning, as
measured by problem scores on self-report measures (the MASC, CDI,
UCLA PTSD index, and YSR).
In a meta-analysis of208 studies utilizing acute psychological stressors,
Dickerson & Kemeny (2004) found that tasks containing both uncontrollable and
social-evaluative elements were associated with the largest cortisol and
adrenocorticotropin hormone changes and the lengthiest recovery times. It is
possible that this biological marker of stress would be concurrent with ratings of
stress. The laboratory stressor utilized in the current study contains both of these
elements. Thus, this investigator predicted that items within the Post Trier Social
Stress Test addressing these elements of the Trier Social Stress Test would be
better predicted by cumulative risk. Specifically, items 4,5, and 8 addressed stress
ratings pertaining to telling your story to the committee, doing mental math in
front of the committee, and having committee members who did not smile or
4. When examining questions 4,5, and 8 individually, cumulative risk scores
(the number of simultaneous risks) at each time period (years 0-2, years 3-4,
6 years, and 12 years) will predict Total PTSST Scores (will rate the Trier
Social Stress Test as more stressful).
5. When examining questions 4, 5, and 8 individually, problems in current
functioning, as measured by higher problem scores on self-report measures
(the MASC, CDI, UCLA PTSD index, and YSR) will predict ratings on the
Total PTSST Scores.
Participants in the current study were fifty-five previously recruited
adolescents aged 17-19, who participated in the Asthma Risk Study since the
prenatal period. Twenty-seven females and 28 males comprised this sample. Fifty
participants (90%) in the current sample reported being White, two participants
(4%) reported Hispanic ethnicity, one participant (2%) reported being Black or
African American, and two reported being Biracial (2%). Approximately 46% of
participants reported having asthma, as goals for recruitment were to match
individuals with/without asthma by gender. Although no current measure of SES
were collected for the current study, information from the larger index sample from
which the current sample was recruited indicated most families were middle class
(Hollingshead I: = 26%, II = 43.3%, III = 22%, IV = 8.7%; Mrazek et al., 1991;
Participants were recruited primarily by phone. A Research Assistant called
potential participants, gave a brief description of the study, and screened them for
eligibility. Because salivary cortisol samples were being taken for another larger
study encompassing this project, participants could not be pregnant, have a Body
Mass Index (BMI) of 35 or more, or have taken oral steroids in the past six months.
Furthermore, female participants were scheduled within five days of the start of
their menstrual cycle. All participants were informed that they would be
compensated $100 for completion of the study.
Letters were also mailed to invite adolescents to participate, so as not to
exclude disorganized families. The letter included contact information for a
Research Assistant, a brief description of the study, and the amount of
compensation available to eligible participants. After potential participants
contacted the Research Assistant, the Research Assistant returned contact and
proceeded as above.
Clinical Rating Scales
Several Clinical Rating Scales were utilized and are individually described
below. These scales were originally created by David Mrazek & Mary Klinnert for
the Asthma Risk Study at National Jewish Medical and Research Center. An
interview-based format was utilized to assess maternal mental health, parenting
risk, and quality of marital relationship. The mother and child were observed
simultaneously during a visit to the familys home, while the mother was given a
semi-structured interview. This allowed the clinician to observe efforts made by
the parent to modulate the child, parental sensitivity, and competence (Mrazek et
Six characteristics of both parents were considered during the parent
interviews, including: (1) the attitude of the parent toward the new infant including
their enthusiasm for their new role as parents; (2) their sensitivity to the needs of
the new infant; (3) their effectiveness in responding to such needs; (4) the parents
strategy for the distribution of parenting responsibilities; (5) evidence of disturbed
emotional adjustment and its potential to impact caring for the infant; (6) adequacy
of the parents plan for employment while providing adequate child care (Mrazek
Various standardized questionnaires were utilized to assess risk when
available. When possible, both clinical ratings and standardized questionnaires
were combined to determine risk in a given area as part of a multimodal method
assessment, a trademark of high-quality research.
Indicators of Risk
A number of the risk factors utilized by Michael Rutter (1979) and in the
Rochester Longitudinal Study (Sameroff et al., 1993, Sameroff Seifer & Zax, 1982)
were utilized as indicators of psychological risk. Due to the retrospective design of
the current study, however, some of the original risk factors from each study were
not able to be included. These alterations/compilations were made in an effort to
incorporate all longitudinal data available for each participant.
Depression Clinical Rating Scales. Clinical rating scales for depression
were available for each participants mother. This rating was made during each
interview following the childs birth. Ratings were made on the mothers report of
mood state. In addition, the mothers affect during the interview was also used as
data. Ratings were made based solely on observed information in the interview; no
inferences about the mothers feelings were used as data. If the mothers mood
state fluctuated over time since the last interview, the lowest point was rated.
Items for the depression clinical rating scale were as follows; A rating of
1 was given for Dysfunctional Depression". This individual was unable to
function on a daily basis and their depression required intervention. A suicide
attempt or detailed suicidal ideation would also have warranted a rating of 1.
Evidence of Severe Depressive Symptoms was rated a 2. This included
interference with daily functioning and/or caretaking of the child and regular
crying. A rating of 3/Moderate Depressive Symptoms, was given when the
parent showed minor, but chronic, symptoms, frequent days of feeling somewhat
down, with a minor decrease in energy level or accomplishments, or significant and
transient problems and/or infrequent episodes of more significant symptoms.
Minor Depressive Symptoms were rated a 4 and point to the existence of short
term, easily resolved blue feelings lasting less than a day, repeated episodes that
are not frequent or regular, and feelings tied to a females menstrual cycle. A 5
indicated No Depressive Symptoms. This means the parent denied any down or
sad feelings. Ratings of 1 or 2 were considered a risk factor for parental
Beck Depression Inventory. The Beck Depression Inventory (BDI; Beck,
Ward, Mendelson, Mock & Erbaugh, 1961) is a 21-item self-report instrument for
measuring the severity of depressive symptoms in adults and adolescents.
Responses are scored from 0 to 3, with a larger number indicating greater severity
of depressive symptoms. Total scores are computed by summing responses for all
21 items. Total scores range from 0 to 63, with higher scores reflecting a greater
severity of depressive symptoms.
This widely used measure of depressive symptoms has good to excellent
reliability and validity for both psychiatric and nonpsychiatric populations. Also,
good internal consistency and test-retest reliabilities have been reported, as well as
concurrence between the BDI and other measures of depressive symptomatology
(Beck, Steer, & Garbin, 1988). Interpretation guidelines for Total Scores on the
BDI are as follows: 0-9 indicates minimal depression, 10-16 indicates mild
depression, 17-29 indicates moderate depression, while scores above 29 signal
severe depression (Groth-Mamat, 2003, p.589). A Total Score > 29 was
considered a risk, if risk was not previously met with information from the clinical
Parenting Risk Clinical Rating Scale. The Parenting Risk Scale
(PRS;Klinnert, Mrazek, & Mrazek, 1994; Mrazek et al., 1995) involved a global
rating from 1-3 intended to capture the clinicians overall sense of how successful
the childs parents were likely to be. Interview questions addressed the couples
attitude toward and preparation for pregnancy and the expected child, their level of
preparation for the expected child, the quality of the couples relationship, the
couples economic situation and potential for appropriate child care arrangements,
psychopathology in either parent (current or historically), the individual parents
family history and related issues, and the likely emotional climate in the home after
babys birth. Clinicians made one rating for the couple based on the interview.
A rating of 1 indicated serious problems such as poor prenatal
planning, severe, untreated psychopathology, history of traumatic separation,
physical abuse in the relationship, or a severely dysfunctional family history. The
clinician viewed this family as needing immediate intervention. Some problems
was rated with a 2. This indicated the clinician felt uncomfortable about the
couple and would have wanted to continue to observe. Dynamic for
problems/tension were present, but were not serious enough to warrant a score of
1. Finally, if no significant problems were noted in the clinical interview, the
couple received a rating of 3. This indicated there was no concerning history, a
stable marital relationship, and no evidence of serious individual psychopathology.
Basically, this describes adequate parenting which includes the range from average
parenting skills to optimal parenting. A rating of 1 was considered a risk factor
for that time period.
Ainsworth Strange Situation Procedure. This gold standard measure of
attachment was administered from 12 to 18 months of age. The procedure involves
exposing an infant to a series of separations and reunions with his/her parent As
this stresses the infants attachment system, his/her behavior in relation to the
parent is rated on four different dimensions: (1) proximity, (2) contact maintenance,
(3) contact resistance, and (4) avoidance. Patterns are reliably classified as secure
(B), insecure avoidant (A), insecure ambivalent (C), or insecure disorganized (D)
(Ainsworth, Blehar, Waters, & Wall, 1978). Indications of insecure attachment
patterns (a rating of A, C, or D) indicated a risk for the respective time period, if
risk was not previously met with information from the clinical scale.
Parental Stress Clinical Ratings. Clinical ratings of maternal stress were
recorded during each interview. During the course of the clinical interview, each
parent was asked about stresses that had occurred since the last interview.
Information was obtained about employment status, economic status, family issues,
extended family, interpersonal issues, and health.
The Stress Scale ranged from 1-5. A score of 1 indicated Extremely
severe stress. Objectively, this indicated the presence of two or more severely
stressful events. In addition, there can be any number of moderate or minimal
events. Severe stress was rated a 2, indicating one severely stressful event and
several lesser ones, or, three or more moderately difficult events. A rating of 3
indicated moderate stress. This would capture two moderately difficult events
with the potential for any number of minimally difficult events. A rating of 4, or
minimal stress, was warranted if the participant experienced one moderately
difficult event, and/or, two or more minimally difficult events. No stress was
given a rating of 5 indicating no moderately difficult events with the maximum
of one minimal stress event. Ratings of 1 or 2 during any of the interviews
were considered a risk for Parental Stress.
Family Inventory of Life Events. The Family Inventory of Life Events
(FILE; McCubbin, Patterson, & Wilson, 1983) is a 171-item self-report measure
designed to record life events and changes experienced by a family in the past six
months. This measure lends itself to quantifying the magnitude of stress
experienced by a family. Items are worded to reflect a significant magnitude of
change that would require adjustment in the regular patterns of interactions
between family members. Such change may be positive or negative. Eight
categories compose the FILE: (1) family development and relationships, (2) family
and extended family relationships, (3) family and work, (4) family management
and decisions, (5) family and health, (6) family and social activities, (7) family and
finances, and (8) family and law. For each change, the family member is asked
whether it was anticipated, the amount of adjustment required (weighted 0-8), and
whether the adjustment continues or is completed (Patterson & McCubbin, 1983).
Preliminary descriptive statistics indicated normally distributed FILE
scores. Thus, individuals whose scores were equal or greater than one standard
deviation above the mean were considered as having risk in this area, if risk was
not previously met with information from the clinical scale.
The Hollingshead Four Factor Index of Social Status. The Hollingshead
Four Factor Index of Social Status is a well-established instrument that explores a
multidimensional approach to SES: occupation, years of schooling, sex and marital
status. This widely used measure of socioeconomic status is computed from
information about education and occupation (Hollingshead, 1975). A
Hollingsheads Four Factor Index Score (I-V) was determined at baseline and when
the child was 72 months of age. Hollingshead classes IV and V consist of semi-
skilled and unskilled laborers, respectively. Categories I, II, and in consist of all
other classes, including skilled craftsmen, clerical and sales workers, minor and
technical professionals, and major business professionals. A rating of IV or V at
the baseline assessment was counted as a risk for prenatal, 0-2 years, and 3-4 years.
At 6 years, new SES scores were calculated. A rating of IV or V was considered a
risk for years 6 and 12, as SES is generally considered stable.
Severe Marital Distress.
Quality of Marital Relationship Scale. The Quality of Marital Relationship
Scale (Klinnert et al., 1992), a clinical rating from 1-5 of the quality of the parents
relationship, was made with emphasis on the resulting emotional environment for
the child. Observations were taken into account with the couples report. A score
of 1 indicated a Destructive Relationship characterized by traumatic
separations, affairs, physical threats, or ongoing abuse. Negative/High Conflict
relationships were scored a 2, and indicated an unpleasant or negative
relationship, with high conflict, great emotional distance, or little emotional
support. A rating of 3 was given to Functioning but Problematic marriages
with a pattern of interaction likely to lead to increased problems. Or, the
relationship pattern was likely to impede functioning for one of the individuals. A
rating of 4 indicated a Mostly Positive marriage, and a rating of 5 indicated
an Excellent marriage with mutual support. A rating of 1 or 2 was
considered a risk factor for Severe Marital Distress.
Dyadic Adjustment Scale. The Dyadic Adjustment Scale (DAS; Spanier,
1976), a 32-item self-report measure was also administered to parents at the time of
each clinical interview. Items on the DAS deal with how partners relate to each
other and they address such issues as agreement, affection, dyadic satisfaction, and
cohesion. The measure has established internal consistency reliability, and
criterion-related validity. The instrument was carefully developed and has been
widely used both for research and in clinical practice. Reliability coefficients for
internal consistency fall between .76 and .96. Furthermore, the scales are shown to
discriminate between married and divorced samples, as well as between distressed
and non-distressed groups of individuals. Construct validity is reported as .86 and
.88, which is unusually high. Four factors emerge on the DAS: (1) Dyadic
Satisfaction; (2) Dyadic Consensus; (3) Dyadic Cohesion; and (4) Affectional
Expression. Indications of problems in the marital relationship on the DAS was
also taken into consideration when determining risk in this area. A total score
below 100 points is indicative of relationship distress. Thus, a score below 100
was considered a risk for the respective time period.
Divorce. Demographic data was available in this area. Indication of
divorce/separation at a given time period was qualified as a risk, if not previously
met by the other forms of measurement mentioned above.
Demographic data. Demographic Data was used to assess this risk factor.
Any minority status or biracial ethnicity/race was considered a risk for all time
Child Psychological Distress.
Temperament. Up to and including year six, temperament was used as an
indicator of the participants psychological distress. For the assessment of child
temperament, Mrazek et al. (1991) designed a global rating to assess and quantify
the degree of difficulty of temperament during the clinical interview. Judgments
were based on maternal report of infant rythmicity and soothing. Ratings on a five-
point scale ranged from 1 very difficult to 5 = very easy. A score of 1
or 2 was considered a risk in this area.
Child Psychological Risk. Beginning at age one, a Child Psychological Risk
(CPR; Klinnert et al., 2001) score was computed for each participant. This risk
score was modeled after the Parenting Risk Scale. The score reflects a 3-point
scale. Scale points are as follows: 3 = no adjustment problems, 2 = possible
adjustment problems", and 1 = definite adjustment problems. A score of 1
indicated risk in this area.
Child Behavior Checklist. The Child Behavior Checklist (CBCL;
Achenbach 1991a) was filled out by parents beginning at two years to obtain data
on a number of behavioral emotional problems and competencies (Achenbach,
1999). The instrument has demonstrated adequate reliability and validity
(Achenbach, 1991a). This 118-item parent report form includes problems that the
parent rates as 0 (not true in the last 6 months), 1 (somewhat or sometimes
true, or 2 (very true or often true).
Several Problem Scales can be computed for the CBCL including:
Withdrawn/Depressed, Somatic Complaints, Anxious/Depressed, Social Problems,
Thought Problems, Attention Problems, Delinquent Behavior, Aggressive
Behavior, and Self-Destructive Behavior (for boys only). The Withdrawn, Somatic
Complaints, and Anxious/Depressed problem scales can then be grouped under the
title Internalizing. Delinquent Behavior and Aggressive Behavior problem scales
are grouped under the heading Externalizing. Such groupings are meant to reflect a
familiar distinction made in childrens behavioral/emotional problems (Achenbach,
1991a). Finally, a Total Problem Score can be computed by summing all problem
scales. Cut point scores for clinical range on the CBCL will be used for
Internalizing, Externalizing, and Total Problem Scores. The cut point is a score of
> 60 (Achenbach, 1991b, p. 127). Thus, scores > 60 on the Internalizing,
Externalizing, or Total Problem Scales of the CBCL, were considered a risk in this
Asthma Status. Information from medical records, participant report, and
parent report was used to determine whether the adolescent had asthma, a chronic
illness. Also, use of corticosteroids was considered when determining if a
participant has asthma. Presence of this chronic illness was considered a risk factor
for the development of psychopathology.
Social Stress Test.
Trier Social Stress Test. All participants were asked to come to the
laboratory between 1:30 and 3:30pm, with a 2:30 start time strongly preferred.
They completed the Trier Social Stress Test (TSST; Kirschbaum, Pirke, &
Hellhammer, 1993). This test was developed as a tool for investigating
psychobiological stress responses in a laboratory setting. The protocol for the
TSST includes a thirty minute resting period upon a participants arrival,
introduction to a task, five minute preparation for the task (in this case a 5 minute
free speech performance of a story stem and mental arithmetic), performance in
front of two unresponsive judges, a microphone, and video recording devices.
If participants finished their prepared story ending in less than five minutes,
the judges responded in a standardized manner. Suggested responses to this
occasion were: You still have time remaining-tell us more about the characters in
your story, Explain what happened when..., Please try and continue until your
five minutes are finished, Tell us more about____________(fill in the blank).
Following this task, the judges asked the participant to serially subtract the
number thirteen from 1023 as fast and as accurately as they could until they were
told to stop. If the participant made an error, judges responded with, Stop, please
start again, or Stop, you have made an error, begin again. If participants made
several mistakes in one minute, judges asked them to subtract the number seven
starting at 868. During both the free speech and mental arithmetic tasks, judges
were instructed not to smile or respond to the participants performance in any way.
After completion of the mental arithmetic task, the judges debriefed the participant.
Then, the participants were escorted to an adjacent room where they responded to
the Post Trier Social Stress Test (PTSST) questionnaire and continued with the
remainder of the study.
Evaluation of Stressful Experience.
Post Trier Social Stress Test. Following the debriefing session, participants
were escorted to an adjacent room where they filled out an evaluation of the Trier
Social Stress Test: the Post Trier Social Stress Test (PTSST) questionnaire. This
ten-item, five-point Likert-style measure addressed questions pertaining to the
amount of stress caused by various aspects of the TSST. Questions covered such
issues as Having to chew on cotton for saliva collection, Waiting for the
procedure to begin, and Telling your story to the committee. Adolescents were
directed to indicate how stressful they found the experience by assigning a number
between 1 and 5 that best describes how much stress they experienced (1 = "Not
at all Stressful, 2 = Somewhat Stressftd, 3 = Stressful, 4 = "Very
Stressful, 5 = Extremely Stressful). A Total PTSST Score was computed by
summing all items.
Indicators of Current Functioning
Self-Report Measures of Current Functioning. Several assessments of
psychological functioning were collected for the current study.
Multidimensional Anxiety Scale for Children. The Multidimensional
Anxiety Scale for Children (MASC; March 1999) is a 39-item 4-point Likert-style
self-report scale designed for individuals aged 8-19. The instrument has gone
through extensive psychometric evaluation (March, 1999). Main and sub-factors
for the MASC are well-established and have been cross-validated in clinical and
population samples (March, 1999). Factors include: (1) physical symptoms;
(tense/restless and somatic/autonomic) (2) Harm Avoidance (anxious coping and
perfectionism); (3) Social Anxiety (humiliation/rejection and public performance
fears); and (4) separation anxiety and are consistent across gender, race, and age
(March, 1999; March, Parker, Sullivan, Stallings, & Conners, 1997). Test-retest
reliability for the MASC falls in the satisfactory to excellent range across age and
gender (March, Sullivan, & Parker, 1999).
Childrens Depression Inventory. The Childrens Depression Inventory
(CDI; Kovacs, 1992) is a self-report questionnaire intended to assess depression in
7 to 17-year-olds. The 27 items composing the CDI consist of three statements
reflecting different levels of depression severity. Each statement is assigned a
value from 0 to 2. A higher number indicates a more clinically severe symptom.
The CDI has shown excellent internal consistency and adequate test-retest
reliability (Saylor, Finch, Spirito, & Bennett, 1984; Smucker, Craighead,
Craighead, & Green, 1986). Also, the instrument has shown criterion-related
validity and sensitivity to Major Depressive Disorder (Carey, Faulstich, Gresham,
Ruggiero, & Enyart, 1987). It can be used to reliably discriminate between
clinically depressed and clinically non-depressed psychiatric patients (Kovacs,
1992) and between clinical and non-referred subjects using CDI factor scores
(Carey, Faulstich, Gresham, Ruggiero, & Enyart, 1987). Unfortunately, this may
not be a valid indicator of depression in the current sample, as some of the
individuals completing the inventory were 18-19 years old at the time they
completed the inventory. This reflected a lapse in the study design and the time
taken to recruit individuals for participation.
UCLA PTSD Index. The UCLA PTSD Index was developed by
researchers and clinicians at the UCLA Trauma Psychiatry Service to screen for
traumatic events and for DSM-IV symptoms of PTSD in school-age children and
adolescents. Three versions are available: The Child Version, Parent Version, and
Adolescent Version. The version used in the current study, the Adolescent
Version, is intended for individuals aged 13 years or older. This 22-item self-report
instrument provides two types of PTSD severity scores: an Overall PTSD Severity
Score and separate PTSD Severity Sub-scores for Criterion B, C, and D symptoms.
Also, by examining the number of symptoms endorsed from each criterion, PTSD
diagnostic information can be determined. Because the UCLA PTSD Index has
only been recently developed, psychometric properties of the instrument have not
been established (Rodriguez, Steinberg, & Pynoos, 1999).
Youth Self-Report The Youth Self-Report (YSR; Achenbach, 1991c) is
part of a family of instruments intended to assess behavioral/emotional problems
and competencies (Achenbach, 1999). This 112-item self-report measure features
statements that describe problem items and items related to social desirability. For
each statement the adolescent is asked to determine how each item describes
him/her in the past six months and rate items from 1 (Not True) to 2 (Very
True or Often True).
Problem Scales that can be computed on the YSR include: Withdrawn,
Somatic Complaints, Anxious/Depressed, Social Problems, Thought Problems,
Attention Problems, Delinquent Behavior, Aggressive Behavior, and Self-
Destructive Behavior (for boys only). The Withdrawn, Somatic Complaints, and
Anxious/Depressed problem scales can then be grouped under the title
Internalizing. Delinquent Behavior and Aggressive Behavior problem scales are
grouped under the heading Externalizing. Such groupings are meant to reflect a
familiar distinction made in childrens behavioral/emotional problems (Achenbach,
1991c). Finally, a Total Problem Score can be computed by summing all problem
The test-retest reliability of YSR self-ratings is satisfactory. Also, evidence
has been presented for the YSRs content validity and ability to distinguish
between matched referred and non-referred youth. However, the absence of
instruments similar to the YSR somewhat limits testing construct validity
Cumulative Risk Scores
Five time periods were assessed for simultaneous risk: Prenatal, 0-2 years,
3-4 years, 6 years, and 12 years. Risks in each area were summed to create a
Cumulative Risk Score for that time period, emphasizing simultaneous risk. For
each category of environmental risk, only one risk could be assigned. However,
evidence from both clinical interviews and self-report measures was utilized to
determine whether risk was met.
For all analyses, cases were excluded pairwise so that cases with complete
data for the pair of variables being correlated were used to compute correlation
coefficients on which regression analyses were based. This was done in an effort to
maximize the number of cases included in each analysis. Tests for significance
Cumulative risk scores for each time period were quite normally distributed
for time periods 0-2 years and 12 years. This was not the case for time periods 3-4
years and 6 years. In terms of categories of risk, the number of individuals in the
high risk category (four or more risks) decreased over time (See Figures 4.1,4.2,
4.3, & 4.4 below).
Distribution of 0-2 Year Cumulative Risk Scores
0 2 4 6 8
Distribution of 3-4 Year Cumulative Risk Scores
0 2 4 6 8
Distribution of 6 Year Cumulative Risk Scores
0 2 4 6 8
Distribution of 12 Year Cumulative Risk Scores
0 2 4 6 8
Nomber of Individnals (N
Regression analyses were performed to test whether cumulative risk scores
predict total scores on the PTSST. Cumulative risk scores were not predictive of
PTSST total scores, nor were they predictive of items 4 or 5 on the PTSST.
However, cumulative risk scores were predictive of PTSST item 8 scores (having
committee members who did not smile or respond) accounting for just 18% of the
variability in responses to this item, F(4, 36) = 3.2, p = .02. Six-year cumulative
risk scores were negatively related to item 8 responses [B = -.38, p = .01], and 12-
year cumulative risk scores were positively related to item 8 [B =.39, p = .02].
Such spurious results might be explained by the poor distribution of six-year
cumulative risk scores.
For analyses related to the predictive value of cumulative risk scores on
current measures of functioning, separate multiple regression analyses were run,
each with cumulative risk scores as predictors and the total scores for the MASC,
CDI, YSR, and UCLA PTSD Index interchanged as dependent variables.
Cumulative risk scores did not predict total scores for any measure of current
Because this population was, for the most part, subclinical, cumulative risk
may not have been related to current functioning, as there was no indication of
psychopathology. Multiple regression analyses were run with the four cumulative
risk scores as predictors and various subscales of measures of current functioning
as the dependent variable(s). When these tests were performed, it was discovered
that cumulative risk scores were predictive of scores on the YSR Somatic Problems
subscale, predicting 21% of the variance, F(4,36) = 3.7, p = .01. Specifically,
twelve-year cumulative risk scores were positively related to scores on the YSR
Somatic Problems subscale [B = 1.08, p = .002].
Again, multiple regression analyses were performed to test whether
problems in current functioning predict total scores on the PTSST. Total scores on
the MASC, CDI, YSR, and UCLA PTSD Index were predictors. These indicators
of current functioning predicted 25% of the variability in PTSST total scores, F(4,
36) = 4.28, p = .006. Although total scores on the MASC were positively related to
PTSST total scores, [B = .34, p = .001], total scores for the CDI, YSR, and UCLA
PTSD Index were not related to total PTSST scores.
When determining whether current functioning predicts responses to item 4
of the PTSST, a stress rating of telling your story to the committee, findings were
not statistically significant.
In examining the predictive quality of current functioning on responses to
item 5 of the PTSST, which captures the stress rating of doing mental math in
front of a committee, regression analyses showed that current functioning
predicted 22% of the variability in responses to item 5, F(4,37) = 3.92, p = .01.
Total scores on the MASC [B = .05, p = .01] were positively related to responses
on item 5. Total scores for die CDI, YSR, and UCLA PTSD Index were not related
to responses on item 5.
For item 8 on the PTSST which captures the stress rating of having
committee members who did not smile or respond, multiple regression analyses
showed that current iunctioning accounted for just 14% of the variability in
responses to item 8, F(4,37) = 2.7, p < .05. Total scores for the MASC appeared to
be positively related to item 8 responses, [B = .05, p = .02). Also, contrary to
hypothesized direction, YSR total problem [B = -.06, p = .04] were negatively
related to item 8 responses. No other measures of current functioning were related
to responses to item 8.
The issue of a subclinical population may have interfered with relationships
between current functioning and PTSST scores as well. To explore this issue
further, Pearson r correlations were run between all subscales of the MASC, YSR,
UCLA PTSD Index and CDI and total scores on the PTSST. Upon examination of
these relationships, it was discovered that both the Physical Symptoms (r = .31, p =
.02) and Social Anxiety (r = .45, p = .001) subscales of the MASC were positively
to PTSST total scores. Furthermore, several YSR subscales were related to PTSST
total scores. The Anxious Depressed (r = .49, p = .000), Social Problems (r = .27, p
< .05), and Withdrawn Subscales (r = .27, p < .05) were each positively related to
PTSST total scores.
This study sought to answer the question of whether cumulative risk and
ratings of laboratory stress were related. In short, the answer to this question is no.
The hypothesis that cumulative risk scores would predict Total PTSST Scores was
not supported. Nor was the hypothesis that cumulative risk scores at each time
period would predict problems in current functioning supported. However, this
may be explained by the fact that the current sample was, for the most part,
Acccording to current measures of psychological functioning, only a few
individuals could be identified as having clinically elevated scores according to the
respective measure. For PTSD, only one individual was probable for a diagnosis
of PTSD (Rodriguez et al. 1999). On the MASC, only two individuals met criteria
(t-score of 70 or more) for inferring clinical problems (March & Parker, 1999). On
the CDI, three individuals met the cutoff of > 19 for screening depressed samples
(Craighead, Smucker, Craighead, Ilardi, 1998). And, as measured by the YSR
cutpoint of > 63 (Achenbach, 1991c) being in the clinical range for development of
behavioral/emotional disorders, only five individuals met criteria (See Figure 4.1
below). When accounting for overlap of individuals, only six individuals met
criteria for clinically elevated scores on self-report measures (See Figure 4.2
Distribution of Clinically Significant Scores on Self-Report Measures
Number of Individ oils
MASC CDI UCLA YSR NONE
However, it is extremely important to consider the fact that these
individuals may not have clinically significant impairment to warrant the diagnosis
of a psychiatric disorder, despite the scores attained on these self-report measures.
That is, individuals were not recruited on the basis of the existence of any
psychological disturbance, and no information was gathered on current psychiatric
diagnoses. So, it is possible that this finding was the result of a lack of individuals
in this cohort who eventually developed some form of psychopathology.
It is interesting that cumulative risk scores predicted answers to ratings of
having committee members who did not smile or respond. Specifically, six-year
scores were negatively related to this item rating while twelve-year scores were
positively related. It is possible that this is evidence for some social anxiety in
individuals who had higher cumulative risk scores at twelve years. It is also
possible that individuals with high risk at an early age found the stressor pale in
comparison to early stressful experiences. However, the interpretations to this
finding were not fully indulged, as the distribution of six-year cumulative risk
scores was poor, increasing probability of error.
Because of the nature of the current sample, additional exploratory analyses
were run for the hypothesis that cumulative risk is related to current functioning.
This was done through examination of subscales of each psychological measure,
which may be more sensitive to differences between individuals who do not meet
criteria for psychopathology. Further examination of the relationship between
cumulative risk scores and subscales of psychological measures revealed that
cumulative risk scores predicted ratings on the YSR Somatic Complaints subscale.
Specifically, twelve-year cumulative risk scores were positively related to this
subscale. Thus, individuals with a greater number of risk factors at age twelve
reported more somatic problems such as headaches, dizziness, and nausea during
participation in the current study.
Chen, Langer, Raphaelson, and Matthews (2004) found that low SES
individuals (who would presumably face more environmental risks) displayed
greater diastolic blood pressure and heart rate activity, with threat interpretations
partially mediating relationships between SES and reactivity. This finding is
somewhat consistent with the relationship found in the current study between
cumulative risk and the Somatic Complaints subscale of the YSR. It is possible
that individuals who are prone to stress have more physiological problems.
However it may also be the case that the reports of somatic complaints are
unfounded. We have no data to suggest the validity of these reports. However, it
would be interesting in the future to examine the health of middle/upper class
adolescents who are faced with several risk factors and determine whether a similar
relationship exists between risk and physiological responses to social stress.
The hypothesis that problem scores in current functioning would predict
ratings on the Total PTSST Scores was supported. Individuals with high total
problem scores on the MASC rated the laboratory stressor as more stressful. This
finding is not surprising considering the social nature of our laboratory stressor.
Specifically, the MASC captures the existence of social anxiety, which would
obviously lead an individual to rate a social stressor as more stressful. Further
exploration of these hypotheses showed positive relationships between the Physical
Symptoms and Social Anxiety subscales of the MASC. Thus, it appears that
individuals who have anxiety related to social situations and report physical arousal
rate the TSST as more threatening. The YSR Anxious Depressed, Social Problems,
and Withdrawn subscales were also related to total PTSST scores. Again, elements
of anxiety and social issues/withdrawal related to the rating of higher stress for the
Such findings are consistent with the cognitive model of anxiety and
depression which postulates that perceptions of events influence individuals
emotions and behaviors (Beck, 1995). It is possible that individuals with
subclinical symptoms of anxiety and depression perceive the laboratory stressor as
more stressful and rate it accordingly. Thus, the very nature of an individuals
perception of events has the potential to affect ratings of a stressor. It is also
interesting that reports of physical arousal were found to be related to ratings of
laboratory stress. Perhaps the sensation of anxiety symptoms was a trigger to rate
the experience as more stressful. It would be enlightening to examine Whether a
biological marker of arousal such as cortisol or cardiovascular functioning is
concurrent with ratings of the social stressor. That is, it is possible that individuals
who are most aware of their physical response to anxiety do not have a large
biological response. Instead, they fear the fear response itself. One recent study
in individuals with social anxiety found that the anxiety experience and perceived
physiological activation may not be as tightly paired with actual physiological
responses than typically thought. Their findings indicated that regardless of
reported severity of social anxiety, the experience was associated with perceived
physiological activation, but not with actual physiological responding (Mauss,
Wilhelm, & Gross, 2004).
Unfortunately, there are several limitations to the current sample. First, the
sample was small for die nature of the current study (N = 55). Furthermore, when
considerations were made for missing data (some individuals did not participate in
various time points of the original longitudinal design), the number of cases was
often smaller. Also, this sample may represent the systematic exclusion of
disorganized families due to the attrition of such families in a longitudinal design.
Furthermore, the original sample from which this sample was recruited was
predominantly white and middle class. Without the pervasive risk contributed by
racism, discrimination, and poverty, these scores may be benign. Finally, the fact
that few, if any, would meet criteria for psychiatric disturbance did not lend itself to
the study of cumulative risk in such a sample.
Another drawback is, due to the retrospective nature of this study, some of
the originally postulated risk factors were not accounted for in die current
cumulative risk scores. Specifically, there was no measure of overcrowding/family
size, paternal criminality, and admission into the care of a local authority, all of
which might have serious detrimental effects. The dual challenge of being as
comprehensive and consistent as possible across time periods was difficult to
The current study varies considerably from the work of previous studies.
Rutter et al. (1977) related cumulative risk to psychiatric disorder as an outcome in
a large prospective epidemiological study. The Rochester Longitudinal Study
(RLS) was prospective in design and examined the effects of the environment on
several hundred children from birth to adolescence (Sameroff et al., 1982,
Sameroff et al., 1993). Unfortunately, due to the predetermined design of the
study, quality of the available sample, and financial constraints, the current study
could not have been designed in a manner consistent with these hallmark
cumulative risk studies.
One caveat in interpreting the results of the study is to be aware that
individuals who do have a psychiatric disorder may, in fact, find laboratory stress
more threatening. The findings here were not necessarily applicable to such a
population as the sample was primarily subclinical and the individuals in this
sample who were at risk did not have a known diagnosis of a psychiatric
disorder, as this information was not available.
Also, this study did not account for protective factors and resilience. A
distinction can be made between the two. Protective factors are considered
moderators of risk and social adversity that enhance positive developmental
outcomes. Resilience is used to describe a buffer that does not eliminate risks, but
allows the individual to deal with them effectively (Rutter, 1987). It is possible
that factors such as these had a protective effect on the individuals in the sample
and prevented the development of psychiatric disorder. Development is influenced
by a wide variety of experiences outside die home along with die family.
Particularly, individuals may be positively influenced by the atmosphere of their
school. Scholastic attainment alone has been found to have a protective effect
(Rutter, 1979). Stable relationships with adults may also lead to better social
adjustment, even if the adult is not a parent (Rutter, 1979). Protective effects can
also exist within the family through grandparents and siblings (Werner, 2000).
Faith and religious beliefs may also provide stability and meaning to the lives of
individuals facing adversity (Werner & Smith, 1989; 1992). Protective factors may
also exist in the community through friends (Werner & Smith, 1992). The
existence of such protective factors in an individuals environment can, in turn,
enhance resilience. Unfortunately, information on these important contributors to
development was not available for in the current study.
There is at least one significant implication for the findings of this study.
Particularly, if individuals who display deficits in current functioning (e.g. high
anxiety, physical arousal) perceive laboratory stressors as more threatening and fear
such a situation, this may have serious effects on the measure of responses to stress
in such a situation, particularly when individuals are grouped on the basis of
another variable. Researchers need to control for current functioning, namely
measures of anxiety, when comparing group responses to laboratory stress.
Otherwise, the participants ratings of stress or changes in biological markers of
stress may be wrongly attributed to a misleading variable. In short, even
subclinical levels of anxiety may mediate ratings of stress or a physiological
response for this stressor.
Future research studies examining the effects of cumulative risk on ratings
of stressful situation should be sure to consider the stressor and its differential
effects on different levels of psychological functioning. It appears that individuals
high in anxiety, particularly social anxiety, tend to rate the laboratory stressor in
this study as more stressful. However, findings may have been different if the
stressor involved less social evaluation and more isolated physical stress.
Achenbach, T. (1991a). Manual for the Child Behavior Checklist and 1991
profiles. Burlington, VT: University ofVermont Department of Psychiatry.
Achenbach, T. (1991b). Integrative guide for the 1991 CBCL/4-18, YSR, and TRF
profiles. Burlington, VT: University of Vermont Department of Psychiatry.
Achenbach, T. (1991c). Manual for the Youth Self-Report and 1991 profiles.
Burlington: VT: University of Vermont Department of Psychiatry.
Achenbach, T. (1999). The Child Behavior Checklist and related
instruments. In M. E. Maruish (Ed.), The use ofpsychological testing for
treatment planning and outcomes assessment (chap. 15, p. 429).
New Jersey: Lawrence Erlbaum Associates, Publishers.
Acker, M. M. & OLeary, S. G. (1996). Inconsistency of mothers feedback and
toddlers misbehavior and negative affect. Journal of Abnormal Child
Psychology, 24(6), 703-714.
Ainsworth, M. D. S., Blehar, M. D., Waters, E., & Wall, S. (1978). Patterns of
attachment: A psychological study of the strange situation. Hillsdale, NJ:
Lawrence Erlbaum & Associates.
Amato, P. R. (1991). Parental divorce and the well-being of children: A meta-
analysis. Psychological Bulletin, 110,26-46.
Beck, A. T., Steer, R. A., & Garbin, G. M. (1988). Psychometric properties of the
Beck Depression Inventory: Twenty years of evaluation. Clinical
Psychology Review, 8,77-100.
Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An
inventory for measuring depression. Archives of General Psychiatry, 4,
Beck, J. S. (1995). Cognitive therapy: Basics and beyond, (p. 14). New York: The
Block, J. H., Block, J. & Gjerde, P. F. (1986). The personality of children prior to
divorce: A prospective study. Child Development, 57, 827-840.
Brady, S. S. & Matthews, K. A. (2002). The effect of socioeconomic status and
ethnicity on adolescents exposure to stressful life events. Journal of
Pediatric Psychology, 27, 575-583.
Carey, M. P., Faulstich, M. E., Gresham, F. M., Riggiero, L., & Enyart, P. (1987).
Childrens Depression Inventory: Construct and discriminant validity across
clinical and nonreferred (control) populations. Journal of Consulting and
Clinical Psychology, 55(5), 755-761.
Carter, A. S., Briggs-Gowan, M. J. & Davis, N. O. (2004). Assessment of young
childrens socio-emotional development and psychopathology: Recent
advances and recommendations for practice. Journal of Child Psychology
and Psychiatry, 45(1), 109-134.
Carter, A.S., Garrity-Rokous, F.E. Chazan-Cohen, R., Little, C. & Briggs-Gowan,
M. J. (2001). Maternal depression and comorbidity: Predicting early
parenting, attachment security, and toddler social-emotional problems and
competencies. Journal of the American Academy of Child and Adolescent
Psychiatry, 40(1), 18-26.
Chakraborty, A. & McKenzie, K. (2002). Does racial discrimination cause mental
illness? British Journal of Psychiatry, 180,475-477.
Chen, E., Langer, D. A., Raphaelson, Y. E., & Matthews, K. A. (2004).
Socioeconomic status and health in adolescents: The role of stress
interpretations. Child Development, 75(4), 1039-1052.
Chen, E. & Matthews, K. A. (2001). Cognitive appraisal biases: An approach to
understanding the relation between socioeconomic status and cardiovascular
medicine. Annals of Behavioral Medicine, 23,101-111.
Chen, E. & Matthews, K. A. (2003). Development of the Cognitive Appraisal and
Understanding of Social Events (CAUSE) videos. Health Psychology, 22,
Clements, A. D. (1992). The incidence of attention deficit-hyperactivity disorder in
children whose mothers experienced extreme psychological stress. Georgia
Psychological Researcher, 91,1-14. (as cited in Huizink, Mulder, &
Compas, B. E. & Boyer, M. C. (2001). Coping and attention: Implications for child
health and pediatric conditions. Journal of Developmental and Behavioral
Pediatrics, 22(5), 323-333.
Compas, B. E., Davis, G. E., Forsythe, C. J., & Wagner, B. M. (1987). Assessment
of major and daily stressful events during adolescence: The Adolescent
Perceived Events Scale. Journal of Consulting and Clinical Psychology, 55,
Craighead, W. E., Smucker, M. R., Craighead, L. W., & Ilardi, S. S. (1998). Factor
analysis of the Childrens Depression Inventory in a community sample.
Psychological Assessment, 10(2), 156-165.
Dickerson, S. S. & Kemeny, M. E. (2004). Acute stressors and cortisol responses:
A theoretical integration and synthesis of laboratory research. Psychological
Bulletin, 130(5), 355-391.
Dodge, K. A., Pettit, G. S., Bates, J. E., & Valente, E. (1995). Social information-
processing patterns partially mediate the effect of early physical abuse on
later conduct problems. Journal of Abnormal Psychology, 104, 632-643.
Downey, G. & Coyne, J. C. (1990). Children of depressed parents: An integrative
review. Psychological Bulletin, 108,50-76.
Emery, R. E. (1982). Interparental conflict and the children of discord and divorce.
Psychological Bulletin, 92,310-330.
Evans, G. W. (2003). A multimethodological analysis of cumulative risk and
allostatic load among rural children. Developmental Psychology, 39,924-
Fergusson, D. M., Horwood, L. J., & Lynskey, M. (1994). The childhoods of
multiple problem adolescents: A 15-year longitudinal study. Journal of
Child Psychology and Psychiatry, 35,1123-1140.
Garcia Coll, C. & Garrido, M. (2000). Minorities in the United States:
Sociocultural context for mental health and developmental
psychopathology. In A. J Sameroff, M. Lewis, & S. M. Miller (Eds.),
Handbook of developmental psychopathology (2nd ed., chap. 10). New
York: Kluwer Academic / Plenum Publishers.
Gelfand, D. M. & Teti, D. M. (1990). The effects of maternal depression on
children. Clinical Psychology Review, 10,329-353.
Graham, P., Rutter, M., & George, S. (1973). Temperamental characteristics as
predictors of behavior disorders in children. American Journal of
Groth-Marnat, G. (2003). Handbook of psychological assessment. Hoboken, NJ:
John Wiley & Sons.
Grych, J. H. & Finicham, F. D. (1993). Childrens appraisals of marital conflict:
Initial investigation of the cognitive contextual framework. Child
Hollingshead, A. B. (1975). Four Factor Index of Social Stratus. New Haven, CT:
Huizink, A. C., Mulder, E. J. H., & Buitelaar, J. K. (2004). Prenatal stress and risk
for psychopathology: Specific effects or induction of general susceptibility?
Psychological Bulletin, 730(1), 115-142.
Kaugars, A. S., Klinnert, M. D., & Bender, B. G. (2004). Family influences on
pediatric asthma. Journal of Pediatric Psychology, 29(7), 475-491.
Kirschbaum, C., Pirke, K. M., & Hellhammer, D. H. (1993). The Trier Social
Stress Test: A tool for investigating psychobiological stress responses in a
laboratory setting. Neuropsychobiology, 28,76-81.
Klinnert, M. D., Mrazek, P, & Mrazek, D. A. (1992). Quality of Marital
Relationship: A clinical rating scale. Psychiatry, 55,132-145.
Klinnert, M. D., Mrazek, D, A., & Mrazek, P. (1994). Early asthma onset: The
interaction between family stressors and adaptive functioning. Psychiatry,
Klinnert, M. D., Nelson, H. S., Price, M. R., Adinoff, A. D., Leung, D.Y. M., &
Mrazek, D. A. (2001). Onset and persistence of childhood asthma:
Predictors from infancy. Pediatrics, 108(A), E69.
Kovacs, M. (1992). Childrens Depression Inventory. New York: Multi-Health
Krieger, N. (2000). Discrimination and health. In L. Berkman & I. Kawachi (Eds.),
Social epidemiology (pp.80-92). Cambridge, MA: Oxford University Press.
Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal, and coping. New York:
Lieb, R., Wittchen, H., Hofler, M., Fuetsch, M., Stein, M. B., & Merikangas, K. R.
(2000). Parental psychopathology, parenting styles, and the risk of social
phobia in offspring. Archives of General Psychiatry, 57, 859-866.
McCubbin, H. L, Patterson, J. W., & Wilson, L. R. (1983). The Family Inventory of
Life Events. Saint Paul: University of Minnesota, Department of Family
McLoyd, V. (1998). Socioeconomic disadvantage and child development.
American Psychologist, 53,185-204.
March, J. S. (1999). The Multidimensional Anxiety Scale for Children. Toronto:
March, J. S. & Parker, J. D. A. (1999). The Multidimensional Anxiety Scale for
Children (MASC). In M. E. Maruish (Ed.), The use of psychological testing
for treatment planning and outcomes assessment (Ch. 10). Mahwah, NJ:
Lawrence Erlbaum Associates.
March, J. S., Parker, J., Sullivan, K., Stallings, P, & Conners, C. (1997). The
Multidimensional Anxiety Scale for Children (MASC): Factor structure,
reliability and validity. Journal of the American Academy of Child and
Adolescent Psychiatry, 36,554-565.
March, J. S., Sullivan, K., & Parker, J. (1999). Test-retest reliability of the
Multidimensional Anxiety Scale for Children. Journal of Anxiety Disorders,
Mathews, A. & MacLeod, C. (1994). Cognitive approaches to emotion and
emotional disorders. Annual Review of Psychology, 45,25-40.
Mauss, I. B., Wilhelm, F. H., & Gross, J. J. (2004). Is there less to social anxiety
than meets the eye? Emotion experience, expression, and bodily
responding. Cognition & Emotion, 18(5), 631-662.
Mrazek, D. A., Klinnert, M. D., Mrazek, P. & Macey, T. (1991). Early asthma
onset: Consideration of parenting issues. Journal of the Academy of Child
and Adolescent Psychiatry, 30(2), 277-282.
Mrazek, D. A., Mrazek, P. & Klinnert, M. (1995). Clinical assessment of parenting.
Journal of the American Academy of Child and Adolescent Psychiatry,
Mrazek, D. A., Schuman, W. B. & Klinnert, M. (1998). Early asthma onset: Risk of
emotional and behavioral difficulties. Journal of Child Psychiatry and
Psychology, 39(2), 247-254.
Patterson, J. M. & McCubbin, H. I. (1983). The impact of family life events and
changes on the health of a chronically ill child. Family Relations, 32,255-
Repetti, R. L., Taylor, S. E., & Seeman, T. E. (2002). Risky families: Family social
environments and the mental and physical health of offspring.
Psychological Bulletin, 128(2), 330-366.
Rodriguez, N., Steinberg, A., & Pynoos, R. S. (1999). UCLA PTSD index for
DSM-IV (revision 1) instrument information: Child version, parent version,
adolescent version. Unpublished manuscript, UCLA Trauma Psychiatry
Rueter, M. A., Scaramella, L., & Wallace, L. E. (1999). First onset of depression or
anxiety predicted by the longitudinal course of internalizing symptoms and
parent-adolescent disagreements. Archives of General Psychiatry, 56(8),
Rutter, M. (1979). Protective factors in childrens responses to stress and
disadvantage. In M. W. Kent & J. E. Rolf (Eds.), Primary prevention of
psychopathology: Social competence in children (Vol. 3,49-74). Hanover,
NH: University Press of New England.
Rutter, M. (1983). Statistical and personal interactions: Facets and perspectives. In
D. Magnusson & V. Allen (Eds. ), Human development: An interactional
perspective (pp. 295-319). New York: Academic Press.
Rutter, M. (1987). Psychosocial resilience and protective mechanism. American
Journal of Orthopsychiatry, 57,316-31.
Rutter, M. (1993). Stress, coping, and development. InN. Garmezy & M. Rutter
(Eds.), Stress, coping, and development in children (pp. 1-41). New York:
Rutter, M. (2000). Psychosocial influences: Critiques, findings, and research needs.
Developmental Psychopathology, 12,375-405.
Rutter, M. (2005). Environmentally mediated risks for psychopathology: Research
strategies and findings. Journal of the American Academy of Child and
Adolescent Psychiatry, 44(\), 3-18.
Rutter, M., Quinton, D., & Yule, B. (1977). Family pathology and disorder in
children. London: Wiley.
Sadock, B. J. & Sadock, V. A. (2003). Kaplan & Sadocks synopsis of psychiatry:
Behavioral sciences, clinical psychiatry (9* ed.). Philadelphia: Lippincott
Williams & Wilkins.
SamerofF, A. J. (1998). Environmental risk factors in infancy. Pediatrics, 102,
SamerofF, A. J., Bartko, W. T., Baldwin, A., Baldwin, C. & Siefer, R. (1998).
Models of development and developmental risk. In C. H. Zeanah, Jr. (Ed.),
Handbook of infant mental health (pp. 3-13). New York: Guilford Press.
SamerofF, A. J. & Chandler, M. J. (1975). Reproductive risk and the continuum of
caretaking casualty. In F. D. Horowitz, M. Hetherington, S. Scarr-
Salapatek, & G. Siegel (Eds.), Review of child development research (V ol.
4, pp. 187-244). Chicago: University of Chicago Press.
SamerofF, A. J. & Fiese, B. H. (2000). Models of development and developmental
risk. In C. H. Zeanah (Ed.) Handbook of infant mental health (2nd ed., pp. 3-
19). New York: The Guilford Press.
SamerofF, A. J., Lewis, M. & Miller S. M. (Eds.) (2000). Handbook of
developmental psychopathology. New York: Kluwer Academic / Plenum
SamerofF, A. J., Seifer, R., Baldwin, A., & Baldwin, C. (1993). Stability of
intelligence from preschool to adolescence: The influence of social and
family risk factors. Child Development, 64, 80-97.
SamerofF, A. J., Seifer, R., & Zax, M. (1982). Early development of children at risk
for emotional disorder. Monographs of the Society for Research in Child
Development, 47 (Serial No. 199). (as cited in SamerofF, 1998).
Saylor, C. F., Finch, A. J., Spirito, A. & Bennett, B. (1984). The Childrens
Depression Inventory: A systematic evaluation of psychometric properties.
Journal of Consulting and Clinical Psychology, 52,955-967.
Seifer, R. (2000). Temperament and goodness of fit. In A. J SamerofF, M. Lewis, &
5. M. Miller (Eds.), Handbook of developmental psychopathology (2nd ed.,
chap. 14). New York: Kluwer Academic / Plenum Publishers.
Smucker, M. R., Craighead, W. E., Craighead, L. W., & Green, B. J. (1986).
Normative and reliability data for the Childrens Depression Inventory.
Journal of Abnormal Child Psychology, 14,25-39.
Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the
quality of marriage and similar dyads. Journal of Marriage and the Family,
Van den Bergh, B. (1990). The influence of maternal emotions during pregnancy
on fetal and neonatal behavior. Pre- and Perinatal Psychology, 5,119-130.
Vaughn, B. E., Bradley, C. F., Joffe, L. S., Seifer, R., & Barglow, P. (1987).
Maternal characteristics measured prenatally are predictive of ratings of
temperamental difficulty on the Carey Infant Temperament Questionnaire.
Developmental Psychology, 23,152-161.
Wagner. B. M. & Compas, B. E. (1990). Gender, instrumentality, and expressivity:
Moderators of adjustment to stress during adolescence. American Journal of
Community Psychology, 18,383-406.
Ward, A. J. (1991). Prenatal stress and childhood psychopathology. Child
Psychiatry and Human Development, 22,97-110.
Werner, E. E. (2000). Protective factors and individual resilience. In J. P. Shonkoff
& S. J. Meisels (Eds.), Handbook of early intervention (pp. 115-132). New
York: Cambridge University Press.
Werner, E. E. & Smith, R. S. (1989). Vulnerable but invincible: A longitudinal
study of resilient children and youth. New York: Adams, Bannister, Cox.
Werner, E. E. & Smith, R. S. (1992). Overcoming the odds: High risk children
from birth to adulthood. Ithaca, NY: Cornell University Press.
Williams, S., Anderson, J., McGee, R., & Silva, P. A. (1989). Risk factors for
behavioral and emotional disorder in preadolescent children. Journal of the
American Academy for Child and Adolescent Psychiatry, 29(3), 413-419.