Effect of Premature Birth on Psycho-biological Processes of Adults

CHAPTER ONE

Little is known about the psychobiological processes of adults, at age 23, who were born prematurely despite the fact that 1 out of every 9 infants is born too early (Atzil, Hendler & Feldman, 2011; CDC, 2015; Mannisto, Vaarasmaki, Sipola – Leppanen, Tikanmaki, Matinolli, Pesonen, Raikkonen, Jarvelin, Hovi, & Kajantie,2015).   Compared to infants born at term, premature infants, face additional obstacles of immature body systems and more neonatal stress, and are at risk for developmental delay, and possible parental overprotection patterns (Clarke, Cooper & Creswell, 2013; Grunau, 2013; Kopp & Rethelyi, 2003; Pinquart, 2014).

Developmentally, in the United States, 23 year olds are expected to achieve independence and form intimate relationships, yet there is limited knowledge about their emotional intelligence and related psychobiological processes while more is known about their stress levels, coping strategies, and emotional disorders (Clarke, et al, 2013; Ingels, Glennie, Lauff, & Wirt, 2012; Pinquart, 2014; Simpson, 2009).

Intrapersonal, interpersonal, adaptability, stress management effectiveness, as well as general emotional health are all involved in reaching adult developmental milestones (Granger & Kivlighan, 2003; Simpson, 2009).   These important young adults milestones require the ability to:  communicate effectively with others, be sensitive to others, maintain emotional self-control, use both knowledge and experiences to cope, manage stress, work and assume responsibilities (Arnett, 2013; Di Fabio & Saklofse, 2014; Simpson, 2009).

The abilities required to meet young adult developmental milestones are captured in definitions of emotional intelligence (EI) and also incorporates the involvement of the brain’s prefrontal cortex executive functioning which further evolves during this time period (Kristensen, Parker, Taylor, Keefer, Kloosterman, & Summerfeldt, 2014; Arnett, 2013; Davis & Humphrey, 2012; Armstrong, Galligan, & Critchley, 2011; Lishner, Swim, Hong & Vitacco, 2011; Aldao, Nolen-Hoeksema & Schweizer, 2010; Stuss, 2009; Tarasuik, Ciorciari & Stough, 2009; Ciarrochi, Deane & Anderson, 2000). The Hypothalamic – Pituitary -Adrenal (HPA) Axis biological response to stress and prefrontal cortex executive functioning abilities are theorized to be involved with emotional intelligence (EI), functioning, stress, coping, and emotional health (Frodl & Stuart, & Pretorius, 2007; Smith & Vale, 2006; Herman, Ostrander, Meuller & Figueiredo, 2005).

Young adults who were born prematurely carry the consequences of their early birth.  As infants, their immature physical systems were responding to ongoing physical pain from neonatal medical procedures, a longer hospitalization and a multitude of environmental stimuli with unstable and ineffective bodily system responses (Grunau, 2002).  The foundations for brain growth and functioning and what will become their usual stress response were being laid at this point in time and will likely influence lifelong development including meeting young adult milestones.

However, there is a need for research examining the mechanisms of stress and its relationship with adult emotional health outcome to assure the well being of young adults and their attainment of developmental milestones (Mannisto, et al., 2015; Simpson, 2009; Tarasuik, et al., 2009). There is minimal research into emotional health and its association with biological processes functioning existing to help in identifying the precise timing and targets of health promoting interventions.

Theoretical Framework

The Developmental Origins of Health and Disease (DOHaD) theory of HPA dysregulation postulates that the stress of prematurity and during early life critical brain growth periods (i.e. formation completes during young adulthood) effects the HPA Axis structure and functioning (Eiland & McEwen, 2012; Sullivan, Hawes, Winchester & Miller, 2008; Barker, 2007; Miller, Chen & Zhou, 2007; De Boo & Harding, 2006; Pujol, Vendrell, Junque, Marti-Vilalta, & Capdevila, 1993).  Stress responses from the HPA axis are adaptive (allostasis) and if occur frequently, particularly with prematurely born infants and their underdeveloped systems, will result in injury or ‘”wear and tear” (allostatic load) on the body (McEwen, 2003 & 2006; McEwen & Seeman, 2006). If the stress is ongoing or chronic than the “wear and tear” on the body will result in illness (McEwen, 2003 & 2006 McEwen & Seeman, 2006). Cortisol is the primary HPA hormone and it can be measured non-invasively and reliably in saliva (Helhammer, Wust, & Kudielka 2009; Turner-Cobb, 2005; Granger & Kivlighan, 2003). When the body is stressed, the HPA axis releases higher levels of cortisol resulting in metabolic imbalances involving blood sugar and insulin, higher blood pressure, faster heart rate, mood changes, thyroid imbalances and weakened immunological responses (Bruyere, 2009).

A normal salivary cortisol stress reactivity response consists of a peak elevation of salivary cortisol concentrations 10 minutes after the cessation of the test stressor with a decrease to pre-stress levels approximately 90 minutes after the start of the test (Kirschbaum, Pirke & Hellhammer, 1993).   In addition, heart rates peak during the protocol’s stressful task and then drops to baseline once the stressor stops (Kirschaum, Pirke, & Hellhammer, 1993). In a recent study, prematurely born age 6 -10 years olds were found to have an exaggerated cortisol response when faced with a social stressor in a reliable laboratory paradigm, and more emotional and memory problems (Quesada, Tristao, Pratesi & Wolf, 2014). Biological stress responses (i.e. cortisol) play a reciprocating role whether the stress is physical or psychological.

Psychological stress (i.e. social evaluation and the perception of uncontrollability in the situation) activates the cortisol response (HPA Axis), which in turn effects the physical systems. Conversely physical stress activation (i.e. pain) of the cortisol system (HPA Axis) is associated with psychological changes in affective and cognitive processes (Smith & Vale, 2006; Dickerson & Kemeny, 2004).

This study uses a standardized, widely used and well-researched laboratory test for inducing moderate psychobiological stress responses called the Trier Social Stress Test (TSST), (Kudielka, Hellhammer & Kirschaum, 2010; Kirschbaum, Pirke, Hellhammer, 1993). In a meta-analysis of 208 laboratory stress studies, the TSST was found to repeatedly induce changes in the concentration levels of cortisol (both serum and saliva), as well as other major HPA Axis endocrines, and to significantly cause an increase in heart rate (Kirschaum, 2010; Kudielka, Hellhammer & Kirschaum, 2010).   Two key components of the TSST’s protocol are well studied and known to be needed to induce a reliable and strong activation of the HPA Axis measured in salivary cortisol.   The components are the psychological stress of the threat of social evaluation, and the perception of uncontrollability in the situation (Kirschbaum, 2010; Dickerson & Kemeny, 2004).   Over fifteen years of TSST research has shown an increase by 50-300% over baseline in endocrine, immunological and cardiovascular parameters (Kirschbaum, 2010).

Young adults born early with immature body systems may not just simply grow out of it and catch up to those born full term (URI, 2011).   Subtle, minor and clear differences in attention, hyperactivity, and emotional and socialization effectiveness have been found during childhood and adolescence (Healy, 2010).   Guided by the DOHaD theory, the developmental challenges of independent living at age 23 years call for greater knowledge about how mechanisms of stress and capacity for response are seen in former premature infants. There is minimal research on 23-year-old outcomes of stress, coping, emotional intelligence, emotional health, stress reactivity responses and the progression of emotional disorders. More knowledge about the relationship of premature birth, the neuroendocrine stress response, self-reported stress, coping and emotional intelligence will expand our understanding of the well being of young adults and their attainment of developmental milestones. (Mannisto, et al., 2015; Simpson, 2009; Tarasuik, et al., 2009).

Purpose

The Developmental Origins of Health and Disease (DOHaD) framework provides the theoretical perspective for the proposed study. DOHaD asserts that early prenatal and neonatal stress seen in HPA axis function affects later health and behavior.   It offers a mechanism to enable the understanding of salient young adult developmental performance abilities.  This is a secondary analysis of a well- characterized cohort of premature infants categorized by neonatal illness into four groups of prematurely born infants (Appendix A Measurements) and one group of full-term infants (for a total of 5 groups) who were assessed at age 23 years in a research protocol, which included measures of stress, coping, emotional intelligence, emotional health and the Trier Social Stress Test.   The purpose of the study is to:

1. Compare the effect of prematurity on stress, coping, emotional intelligence and emotional health at age 23 years.

2. Examine the neuroendocrine functioning of the stress response and the stress recovery period at age 23 years.

3. Examine emotional health with HPA

Axis stress reactivity responses accommodating for prematurity and gender.   The knowledge gained from this study can help to inform how neonatal stress of prematurity effects young adult coping, emotional intelligence, physiological responses, developmental milestones and emotional health at age 23 years in a sample at risk for HPA dysfunction. The results can help to identify who may be at risk, the role of the neuroendocrine systems as an underlying mechanism, and suggest clinical interventions to be taken to avoid risk factors and promote future adult health outcomes (Rice, 2012; Kirschibaum. 2010; Sullivan, 2008).

Aims and Hypotheses

In a sample of young adults at age 23, former premature infants with a wide variation in diagnoses of neonatal illnesses and a full term group, the aims of the study, with related hypotheses are:

Aim 1. Compare the effect of prematurity on stress, coping, emotional intelligence and emotional health.

Hypothesis 1. Higher self-reported stress scores, higher use of avoidance coping types, lower emotional intelligence scores and more emotional health disorders will be found for the adults at age 23 years born prematurely compared to the term-born adults.

Aim 2. Compare the salivary cortisol response to social stress between premature and term-born infants using stress paradigm of the Trier Social Stress Test (TSST).

Hypothesis 2. Adults at age 23 who were born prematurely will have a prolonged stress recovery period of the TSST.

Aim 3. Examine the relationship between effect of emotional health and on the stress recovery period of the TSST measured in salivary cortisol.

Hypothesis 3. The stress recovery period for adults at age 23 years with emotional health problems will be prolonged compared to adults without emotional health problems when prematurity is controlled.

Summary

This study used a well-characterized, longitudinal sample of preterm and full term born infants who have been followed from birth in a series of research studies. The study used neonatal data and self-report of stress, coping and emotional intelligence, clinical diagnosis of emotional health, and neuroendocrine function during a social stress paradigm at age 23 years in a secondary analysis. The study is congruent with the original, larger, study using the same theoretical framework and relevant variables to examine self-reported stress, coping, emotional intelligence and emotional health with the primary biomarker of HPA axis system, cortisol. The overall aim of the study is to examine self-reports of important emerging adult independent function with neuroendocrine activity in the well-standardized social stress test, the TSST.   The brain areas most affected by stress are the same areas involved in adapting to stress and coping effectively (Compas, 2006). Researching these integrated psychobiological processes through multiple analyses will lead to further understanding of how stress effects young adults emotional development (Compas, 2006). This study has the potential to add relevant knowledge about salient developmental characteristics, elements and competencies. The following chapter (2) provides the scientific literature in support of this study. These include the DOHaD theoretical framework and developmental milestones at age 23 years old of the emerging adult who was prematurely born. The use of secondary longitudinal data for analysis will also be addressed.

CHAPTER TWO

Theoretical Background and Related Literature

In this chapter, selected theoretical perspectives for this study are delineated, specifically the Developmental Origins of Health and Disease (DOHaD), neuroendocrine function of the HPA Axis, and prematurity. Age appropriate young adult development at age 23 years will be described in relation to concepts of stress, coping, emotional intelligence, and emotional health. Prematurity effects compared to the full term born for 23 year olds will be understood within a disease developmental theory integrating both biological and psychosocial aspects.

The discussion will begin with a historical review of the DOHaD theory and the contributions up until present time. Included is a review of the construct development of emotional intelligence and the relationship of DOHaD to stress, coping and emotional health.

Theoretical Perspectives

The relatively new Developmental Origins of Health and Disease (DOHaD) theory has gained scientific acceptance within the last thirty years (Wandhwa, Buss, Entringer & Swanson, 2009). DOHaD postulates that fetal and neonatal stressors affect the neurological and endocrine systems adaptive responses, specifically the hypothalamic-pituitary-adrenal (HPA) axis, which is essential for physical and psychological growth (Barker & Thornburg, 2013; Carpenter, Gawuga, Tyrka, Lee, Anderson, & Price, 2010; Wandhwa, et al., 2009; McEwen, 2003; Sapolosy, 2001).

The major stress pathway of the neurological system is the HPA Axis (Figure 1) which releases cortisol. Cortisol, the most dominant stress hormone that crosses the blood – brain barrier, has the important function of adapting the body to both physical and emotional stress responses (Bruyere, 2009).   Additionally cortisol is vital in the regulation of blood vessel tone, the inflammatory response, stimulation of glucose production, insulin, and metabolism (Bruyere, 2009). HPA functioning is altered by stress during structural growth periods resulting in permanent programming of early life stress responses that contribute to disease formation later in life (Sullivan, Hawes, Winchester & Miller, 2008).

Figure 1 Hypothalamic-Pituitary-Adrenal Axis

Figure 1. Reprint by permission from Worth Publishers. “An Introduction to Brain and Behavior (5th Ed.)” by Bryan Kolb and Ian Q. Whishaw, 2016. Copyright 2016 by Worth Publishers. From: AN INTRODUCTION TO BRAIN AND BEHAVIOR 5E, by Bryan Kolb, et al, Copyright 2016 by Worth Publishers. Used by Permission of the publisher.

Premature birth (<37 weeks gestation) occurs at a rate of one out of every nine births, is the leading cause of infant deaths and long-term neurological disabilities in children (CDC, 2014). The normal duration of pregnancy is 9 months (280 days) with full term birth occurring at 40 weeks gestation (Taber’s, 2009). Preterm birth, either naturally or by cesarean section, is “arbitrarily defined as before 37 weeks” (WHO, 2015; Johansson & Cnattigius, 2010).   Preterm birth can be further subdivided into:  moderately premature (32-<37 weeks), very premature (28-32 weeks), and extremely premature (<28 weeks), (WHO, 2015). The characteristics of low birth weight and rates of fetal growth has also been used to define prematurity.

In preterm research studies, the combination of gestation weeks and birth weights are used to avoid any misclassification especially with infants who have growth restrictions (Johansson & Cnattigius, 2010). Low birth weight is <2500g (5 lbs. & 8 oz.), very low birth weight is <1500g (3 lbs. & 4 oz.), and extremely low birth weight is  <1000g (2 lbs. & 3 oz.), (Johansson & Cnattigius, 2010).

Surviving preterm born infants may have intellectual disabilities, neurological problems, respiratory, visual, hearing and digestive problems (CDC, 2014; Martin & Osterman, 2013). Premature infants have experienced prenatal stress, often from health risk factors in the mother and postnatal stress from months–long intensive care in the Neonatal Intensive Care Unit (NICU).   These early stress experiences evoke broad biological responses in the underdeveloped premature infant’s neurophysiological systems including brain growth occurring during uterine development (Phillips, 2001) and peak growth spurts well into the mid-twenties of age (Epstein, 1986).   Thus, the premature infant’s neuroendocrine system is frequently activated leading to the inductive development and empirical evidence in support of DOHaD theory.

Historical Evidence and Perspectives DOHaD evolved from Barker’s original “Fetal Origins Hypothesis” which originated inductively from epidemiological studies. (Barker, 1990; 2004, 2007; Barker & Thornburg, 2013; Wandhwa, et al., 2009; Gluckman & Hanson, 2007; Reynolds, 2007; Hofman, Regan & Cutfield, 2006).  During this time period multiple retrospective mortality and epidemiological studies from different countries showed evidence that adult height and geographical differences were related to infant mortality caused by heart disease (Barker, 1990, Barker, Erickson, Forsen, & Osmond, 2002; Phillips, 2001).  An influential study, consisting of 499 people at age 50 born in England, revealed their current blood pressure measurements and hypertensive risk factors were strongly related to the measurements of their hospital-recorded placenta and birth weights (Barker & Osmond, 1986; Barker, 1990).   A lack of evidence was found for the role of some commonly involved environmental variables in heart disease, such as a high fat diet, and this finding prompted an alternative hypothesis (Barker, 1995, 2007; Barker & Osmond, 1983).   As a result of this landmark study, a paradigm shift representing a new conceptualization of disease causation occurred (Barker, 2007).   This shift in scientific thinking about adult diseases, which was defined as degenerative in nature and viewed as a result of gene and environmental interactions, occurred and resulted in the inclusion of biological programming during fetal and infant life (Barker, 1990) as a plausible explanation (Gordis, 2009).

Barker’s (1995) initial assumption was:  “ fetal under-nutrition in middle to late  gestation, which leads to disproportionate fetal growth, programs later coronary heart disease (p 171)” lead to further studies from this hypothesis.  Later Barker (2004) refined this to the hypothesis:

“Cardiovascular disease and type 2 diabetes originate through developmental plasticity, in response to undernutrition. As it is now known that growth during infancy and early childhood is also linked to later disease “developmental origins hypothesis’ is now preferred (p. 114).” 

Barker, defined a process of developmental plasticity as: “ a critical period when a system is plastic and sensitive to the environment, followed by loss of plasticity and a fixed functional capacity” (Barker, 2007, p 415).  According to Barker, developmental plasticity also has three features:

1. The response will depend on the nature of the environmental cue.

2. There are critical time periods for different systems when changes will occur and these changes may be temporary or permanent.

3. Duration of these critical time periods are different depending on the structure with the brain periods longer. These changes can be gross, substructure or functional.

The imbalance of fetal nutrients and oxygen was thought to result in an alteration of not only metabolic yet endocrine functioning leading to smaller birth weights and a variety of adult pathologies (Drake, Tang, & Nyirenda, 2007; Phillips, 2007; De Boo & Harding, 2006, Gibson, Carney & Wales, 2006). Despite the scientific acceptance of this explanation, methodological issues surrounding the use of body measurement, birth weight or gestation age, and studies designed without a well-characterized cohort utilizing prenatal and adult health outcomes, as well as observational and prospective designs, added to concerns about confounding variables (socioeconomic status, diet, cigarette smoking, physical exercise and selection bias), statistical effect sizes (attrition and statistical over adjustments) and publication bias (Skogen & Overland, 2012; Erickson, 2006; Godfrey, 2006). A 2003 meta-analysis study addressing publication bias relating to low birth weight and higher blood pressure found a weaker association than initially determined yet maintained support for the fetal origins hypothesis (Skogen & Overland, 2012). Despite a weaker association found, while addressing some of the common confounding variables as alternative explanations, better research methodologies resulted and improved the replication of findings (Gordis, 2009).

Further studies from a variety of countries designed to control for confounding variables supported the association between low birth weights as a fetal antecedent to diseases (Barker & Bagby, 2005; Vohr, Wright, Dusick, Mele, Verter, Steichen, Simon, Wilson, Broyles, Bauer, Delaney-Black, Yolton, Fleisher, Papile & Kaplan, 2000). At the same time debates occurred focusing on the idea that the only important applicable time period for DOHaD was during pregnancy and the theory was useful in explaining only cardiovascular diseases (Gluckman & Hanson, 2006; Godfrey, 2006).   Indeed, DOHaD flourished in explaining cardiovascular risk (Bryan & Hindmarsh, 2006). Later, researchers significantly correlated low birth weight with increased risk in a number of diseases that are part of the metabolic syndrome (Hofman, Regan & Cutfield, 2006) such as truncal (middle body) obesity, hypercholesterolemia, atherosclerosis, coronary artery disease, insulin resistant diabetes mellitus, cardiomyopathy, congestive heart failure, autoimmune disease, anxiety, depression, chronic pain and headache (Fricchione, 2015). Replicable cross sectional studies of relationships between disease formations with metabolic illnesses supported further study of the HPA axis associations.

The results from a retrospective longitudinal study of Helsinki, Finland male adults, gave credence to the findings that smaller infants have a higher rate of cardiovascular disease for men in adulthood (Ericksson, Forsen, Tuomilehto, Osmond & Barker, 2001, Godfrey, 2006).  The sample size consisted of 4,630 men born in Helsinki (1934-44) and utilized child welfare clinic medical health records with multiple time points of childhood through adult weight recordings, height measurements and hospital admissions for coronary heart disease (Ericksson, et al., 2001). Overall, low birth weight was associated with heart disease, low weight gain was associated with an increased risk of heart disease and rapid weight gain after age 6 was associated with further risk (Ericksson, et al., 2001).   As a result of this study, determining what changes occurred in prenatal growth and those that occurred later lead to focusing on the interactions of both prenatal and postnatal environments in the development of adulthood cardiovascular diseases (Ericksson, 2006; Godfrey, 2006).

Concurrently animal researchers showed that exposure of rats during pregnancy and after birthing, along with their offspring, to a variety of stressors resulted in elevated stress-induced cortisol levels in the off-spring and disease development (Phillips, 2001).   This study added a “dose-response” relationship, or intensity and timing of the exposure of stressors to the DOHaD research literature. Animal studies, pinpointed the HPA Axis response to stressors introduced during critical times of rat brain growth, that occurred after birth and produced permanent changes in the animals HPA Axis response (Matthew, 2002; Phillips, 2001).   As mentioned earlier, in humans, critical brain growth occurs during uterine development (Phillips, 2001) and continues with peak growth spurts well into the mid-twenties of age (Epstein, 1986).

Sterling and Eyer in 1988 coined the word “allostasis” (Sterling & Schulkin, 2004) based on research with monkeys while studying high blood pressure.   Allostasis was now a new paradigm to explain arousal pathology and replaced homeostasis conceptually.   Allostasis involves regulation by: varying parameters and variations in anticipated demands.  A new core assumption now was that physiology is sensitive to social relations.   Allostasis also depended on higher-level brain functioning, other then basic physiological automatic responses, and involved prefrontal cortex regulation.   Anticipatory regulation for anxiety and satisfaction was found to rely on the prefrontal cortex through neuronal mechanisms.

McEwen, in 1989, further developed these principles through “allostatic load” and is credited with the advancement of the theory by publishing research findings related to human autonomic, central nervous system, endocrine and immune system activity (Sterling & Schulkin, 2004).    McEwen (2006) implicated stress to an event as an individual biological response factor in the development of a disease.   In addition to acute stress events, McEwen defined the effects of general “wear and tear” (p 367) on the body as allostatic load that targets the HPA Axis, releasing an end product of cortisol and can lead to the development of adulthood diseases (McEwen, 2003 & 2006).

The main hypothesis of the DOHaD theory, involves one sensitive brain area of prenatal and postnatal development occurring in adulthood disease development by resetting the glucocorticoid endocrines, which is the Hypothalamic-Pituitary –Adrenal (HPA) Axis (Sloboda, Newnham, Moss and Challis, 2006).   McEwen (2008) further delineated this dimension of the theory by postulating that stress hormones have a central effect in health and disease by providing protective, damaging and mediating effects.

These mediating stress effects can be from a physical, psychological, emotional, cognitive, intellectual, major life events, environmental or social/caring interactions.

Biologically, individuals can either adapt to acute stress (allostasis) or become overloaded (allostatic load) with chronic stress resulting in pathophysiological changes (McEwen & Seeman, 2006).  See Figure 2.

Figure 2 Stress Response and Development of Allostatic Load

Figure 2. Reprinted by permission from Macmillan Publishers Ltd; Neuropsychopharmacology, 2000, by B., McEwen, Allostasis and allostatic load: Implications for neuropsychopharmacology, Neuropsychopharmacology, 22, 108-124. Copyright 2000 by Nature Publishing Group.

The brain is viewed as not only the controller of the stress response yet conversely as the target (Rubinow, 2006). The HPA Axis as programmed is effected by the totality of lifelong stressors (cumulative risk) and negative effects on the brain and body (allostasis/allostatic load) leading to the long-term consequences of adulthood disease (Manzanares, Monseny, Ortega, Montalvo, Franch, Gutierrez-Zotes, Reynolds, Walker, Vilella & Labad, 2014; McEwen, 2012, Sullivan, et al., 2008). In other words, the adaptability to a stressor or anxiety rather than the initial reaction will predict long-term outcomes and allostatic load becomes the unifying concept between cumulative risk and HPA dysregulation (Manzanares, et al, 2014; McEwen, 2012, Sullivan, et al., 2008).

McEwen did initially base his theory on the idea of homeostasis, which conceptually is a bodily system that is stable and unchanging (Dictionary.com, 2015) and postulated there is an optimal level and ideal set point (McEwen, 2004).   This explanation evolved to include the idea of variation (allo) of levels and set points achieving a balance in the total system (Dictionary.com, 2015).   McEwen points out:

Homeostasis is about adjusting this level while allostasis is about the brain coordinating body-wide changes to achieve stability through change (McEwen, 1998 & 2004). Thus adaptation in a central concept of the theory.   Stressors, according to McEwen result in experiences that are either acute or chronic.   Acute stress is the “fight or flight” response or those responses resulting from major life events. Chronic stress is defined as the accumulation of minor and daily stresses.

There are four types of allostatic load (Figure 3):  normal, repeated, lack of adaptation and inadequate (McEwen, 2000 & 2007).

Figure 3 Four Types of Allostatic Load

“Four types of allostatic load are illustrated. The top panel illustrates the normal allostatic response, in which a response is initiated by a stressor, sustained for an appropriate interval, and then turned off. The remaining panels illustrate four conditions that lead to allostatic load: 1) Repeated “hits” from multiple novel stressors; 2) Lack of adaptation; 3) Prolonged response due to delayed shut down; and 4) inadequate response that leads to compensatory hyperactivity of other mediators, e.g., inadequate secretion of glucocorticoid, resulting in increased levels of cytokines that are normally counter-regulated by glucocorticoids). Figure drawn by Dr. Firdaus Dhabhar, Rockefeller University.” 

Figure 3. Reprinted by permission from Macmillan Publishers Ltd; Neuropsychopharmacology, 2000, by B., McEwen, Allostasis and allostatic load: Implications for neuropsychopharmacology, Neuropsychopharmacology, 22, 108-124. Copyright 2000 by Nature Publishing Group.

Endocrine and metabolic responses protect the body from allostatic load in the short term through homeostatic adaptation called allostasis.   Chronic stressors or allostatic load, whether physical, psychological or a combination, result in structural brain changes that effect our physiological and behavioral responses and lead to the development of adulthood diseases (Ewen, 2003 & 2006).

The stress experience or the stress response that the individual has to a potential stressor, is the focal point (McEwen, 1998) and added to “Barker hypothesis” of prenatal stage development, the future time determinants of adulthood diseases (Skogen & Overland, 2012).   Stress as defined by McEwen is a “state of real or perceived threat to homeostasis” and stressors are “aversive stimuli” while “maintaining homeostasis through activation of complex responses involving the endocrine, nervous and immune system” is the stress response (McEwen, 2006).

Luthar, Cicchetti and Becker, 2000, are clearer in their definition of the stress response, as a dynamic equilibrium, meaning an ability to sway and not as a fixed or static state. When a good adjustment is achieved across different domains of the stress response, in the face of significant adversity, then “resiliency” is achieved (Luthar, Cicchetti & Becker, 2000; Ficchione, 2015).   Thus, the capacity to maintain allostasis while challenged by mental and physical aversive stimuli to well being, constitutes resiliency (Ficchione, 2015).   One Mind Body Medicine equation (hypothesis) is formulated as (Ficchione, 2015):

Stress (Allostatic Loading)  = Selective Vulnerability: Propensity to physical and Resiliency Factors mental illness Ficchione (2015) defines resiliency factors as: “relaxation response, mindfulness; social support/prosociality; cognitive skills; positive psychology; spirituality; exercise; nutrition; healthy habits”.

According to McEwen (1998) there are two factors that determine the individuals stress response:  how the situation is perceived and the individuals’ general state of health (determined by genetics, behavior and lifestyle choice), (See Figure 2).

Matthews (2002) added that the timing and intensity of the aversive event and/or an intervention also effects HPA axis development and functioning.   This focus on later stress experiences added environmental triggers to critical or sensitive growth time periods as a second possible causal pathway to disease suggesting the involvement of more than one critical time period (Skogen & Overlad, 2012).   Stressors, occurring prenatally result in adaptive changes within the HPA-Axis, become permanently programmed, and impact health during adult life (Reynolds, 2013; Sullivan, Hawes,

Winchester & Miller, 2008; Barker, 2007) while later environmental triggers and responses add to the allostatic load depending on coordination with sensitive growth time periods (Skogen & Overland, 2012).

The body’s stress response helps individuals adapt to a problem and marshal the resources to respond which includes releasing response coordinating hormones. A unifying and central relationship is if the stress response is activated too frequently or under utilized then the stress-response itself can be more harmful than the stressor and this concept is called allostatic load (McEwen, 1998 & 2004).  There has been controversy over the labeling of this phenomenon yet the underlying concept has not been challenged. McEwen (2004) does identify features of some stress processes that do not change in order to help and adds those processes that do vary in the context of life cycles, individual experience and responses to the physical and social environment (McEwen & Wingfield, 2010). The challenge, according to McEwen, in the definition is that allostasis adds to homeostasis a focus on how individuals have access to bodily resources to respond to problems with the environment (McEwen, 2004).

Welberg and Seckl, 2001, found that stress during pregnancy could permanently alter behavioral and/or physiological reactivity to stressors (Sullivan, et al, 2008).

The authors extensively reviewed available research of epidemiological, animal biological, human biological, anxiety, cognition, neural mechanisms, under-nutrition, interactions with postnatal environments and glucocorticoid studies. In terms of HPA function, the evidence showed:

“Birth weight correlates closely with HPA measures from infancy (206), through adolescence and young adulthood (207) to old age (208). These data suggest that low birth weight associates with both increased basal and ACTH-stimulated cortisol levels (207, 209). Taken as a whole, these findings are compatible with the hypothesis that fetal overexposure to glucocorticoids whether exogenous DEX or endogenous cortisol may underlie at least in part the connection between the prenatal environment and adult stress-related and behavioral disorders (Welberg & Secki, 2001, p 123).”

As this evidence became available another shift in thinking about adulthood disease formation occurred and became widely accepted by the scientific community (Skogen & Overland, 2012; Salonen, Kajantie, Osmond, Forsen, Yliharsila, Paile-Hyvarinen, Arker & Eriksson, 2011; Gluckman & Hanson, 2006; Godfrey, 2006).

The scientific community was now in consensus that relevant life periods are on a continuum that includes during pregnancy, infancy and throughout the life span (Gluckman & Hanson, 2006; Godfrey, 2006).    The focus of the theory was now based on two main assumptions.   The first assumption is early life events that occur during periods of critical biological growth partially determine future adulthood disease development and secondly, this has implications for both disease development and promotion of health (Gluckman & Hanson, 2006; Godfrey, 2006).  DOHaD theory research then branched into three major areas of interest: 1.  Maternal, fetal and postnatal nutrition, 2.  Preterm birth and, 3.  Epigenetics (or gene modification), (Wadhwa, et al., 2009; Waterland & Michels, 2007).

DOHaD theory has been applied to a variety of diseases and health concerns: behavioral, cancer, cognitive, diabetes, metabolic, muscular, neurological, psychological and respiratory (Barker & Thornburg, 2013; Gluckman & Hanson, 2007; Hofman, Regan & Cutfield, 2006; Barker, 2005 & 2004). The theory is widely used in behavioral medicine and specifically with interventions directed at reducing stress responses (Benson, 2015).  Researchers are also exploring the effects of stressors, as measured by cortisol levels and magnetic resonance imaging, with major psychiatric disorders such as post-traumatic stress disorders, anxiety responses, stress responses, cognitive functioning, eating disorders, childhood disorders, personality, long-term effects of child abuse, psychosis and addictions (Nosarti, Murray & Hack, 2010). Indeed, anxiety, an autonomic nervous system response triggered by HPA functioning, is a general symptom of stress and co-occurs with other psychiatric disorders, especially depressive disorders (APA, 2000).

Theory Analysis

DOHaD a relatively new theory addressing the origins of adulthood disease is widely accepted, utilized in clinical interventions and research studies.   Originating from epidemiological study results DOHaD has developed into a major disease causation theory.   The original theory was intended to explain one aspect of disease causality and to be applicable to multiple diseases including those “future entitles yet unknown” (Barker, 2007).   Extending the original Fetal Origins Hypothesis, beyond the initial hypothesis of environmental influences during pregnancy has an effect on later development, expanded the perspective beyond biological determinism (Skogen & Overland, 2013).   The addition of stress responses through further refinement of the HPA Axis dysregulation hypothesis, multiple critical time periods and life span influences took into consideration the role of other causal issues.   In addition to physical stressors, an individual’s perception of stress as threatening or uncontrollable has been shown to activate the HPA axis as well as coping styles choices affects on later life stress-related disease development (Figure 4). The continued development along this thinking moved DOHaD to a fuller life course perspective with multiple causal factors including emotional health (individual perceptions of stress interpreted as threatening or not) environmental (such as parenting) and for some the addition of the biological bases of mammalian evolutionary attachment (social influences) perspectives (Skogen & Overland, 2013; Fricchione, 2011).

DOHaD theory defines a plausible biological temporal relationship between disease formation and the role of the HPA axis. Research results show an association, of a dose-response relationship of intensity and timing of stressors, and replication of Figure 4 findings in specificity to stress-related illness, although the majority of research in the United States, has been at discrete time periods in a person or population’s life.

The Cardiovascular Toll of Stress (Emotional and Physical Stressors, HPA Axis & Health)

Figure 4. Reproduced with permission of Lancet Publishing Group; Brotman, D.J., Golden, S. H., & Wittstein, I.S. (2007) The cardiovascular toll of stress, The Lancet, 370, 1089a1100.

DOHaD theory states the necessary condition of structural and functional changes prenatally and whether any of this is irreversible remains to be seen (Skogen & Overland, 2012). Some common confounding variables as alternative explanations have been addressed and the national government (National Institute of Health) has prioritized the use of longitudinal studies to address the complexities of what factors (NIHR, 2015), may or may not be sufficient or necessary for disorders.

Literature recommendations for further research focus on utilizing regression modeling statistical strategies to address:  the association between the two variables of early exposure and adult outcomes, intermediate exposures, the interaction between the early exposure and intermediate variables and to what degree the intermediate variable is related to disease outcome (Skogen & Overland, 2012).   Research into the effects of stress experiences on HPA Axis development, function and dysregulation also requires addressing the roles of birth term, gender and social inequalities as confounding moderating variables (Matthew, 2002; Sapolsky, 2009).   The renewed interest in DOHaD theory is leading researchers into areas of study that promise to identify risks and protective mechanisms; locate periods of transitions into pathology, develop preventive and possibly corrective interventions to intervene in the progression of disease pathologies over the course of a lifetime (McEwen & Wingfield, 2002; Sullivan, Hawes, Winchester & Miller, 2008; Ben-Sholmo & Kuh, 2002).

Socioeconomic Status (SES)

Many socioeconomic status (SES) factors are associated with low birth weight, coronary heart disease, under nutrition, low literacy rates and health disparities (Senterfitt, Long, Shih, & Teutsch 2013; Baber, Muzaffer, Khan, Imdad, 2010).

Social and economic factors are considered the largest single predictor of health outcomes and influencer of health behaviors (Senterfitt, et al., 2013).   Disparities between countries in preterm birth weights have been partly explained by differences in SES (Johansson & Cnattigus, 2010).   Multiple studies have found that the lower the social, education, and economic position the higher the unhealthy behaviors (i.e., smoking, physical inactivity) and inability to engage in healthy behaviors (Senterfitt, et al., 2013).

The World Health Organization (WHO, 2015) includes determinants of SES consisting of:  the physical environment, the person’s individual characteristics and behaviors (including how they “deal with life’s stresses and challenge, (p 1)”, social support networks, genetics, available health services and gender.   Additionally, a WHO (2003) sponsored study, found middle-class office workers and lower ranking staff has more disease and die earlier than higher positioned workers. The WHO report (Wilkinson & Marmot, 2003) focuses on ten main areas of what is known:

1. Life expectancy and shorter life spans occur further down the social ladder.

2. Stressful circumstances lead to worry, anxiety, inability to cope that is damaging to health.

3. Foundations are laid in childhood: “slow growth and poor emotional support raise the lifetime risk of poor physical health and reduce physical, cognitive and emotional function in adulthood, (p 14)”.

4. Poor life quality shortens lives.

5. Stressful workplaces increase the risk of disease.

6. Health risks of unemployment increases the rates of premature death.

7. Supportive relationships contribute to health.

8. Individuals addictions to alcohol, drugs and tobacco numb difficult conditions and lead to downward mobility.

9. Healthy food is a political issue.

10. People’s dependence on cars has increased resulting in less walking and social contact and more traffic accidents and air pollution.

One of the earliest criticisms of the interpretations from DOHaD studies is the the confounder of SES could explain results used to support the theory (James, Nelson, Ralph, & Leather, 1997; Skogen & Overland, 2012).   James, et al, (1997) found lower socio- economic groups have more premature and low weight births, more illnesses, more risk factors and less nutritional diets.   The DOHaD theory emphasizes that chronic exposure to stress mediators of the HPA Axis and the sympathetic nervous system effects multiple organs resulting in disease (Dowd, Simanek & Alello, 2009) although SES confounding variables in earlier studies were not adequately controlled or interpreted.

Allostatic load has been used to explain part of the association between SES, health and disease.  Lupien, King, Meaney, & McEwen (2001), as one goal of three, explored the possibility that because lower SES status involved higher stress and fewer coping resources then morning salivary cortisol levels would differ from other SES groups.   The cross-sectional experimental design study, utilized 307 children from a school setting, with 6 age groups (6-16 years old) and two categories (low and high) of SES.   Overall findings showed that lower SES in children related to higher cortisol levels with the impact of SES on cortisol absent after transition to high school.

The authors identified four possible social explanatory factors of:  changing status, influence of peers, influence of youth culture, and resilience.

In contrast a systematic review of the literature, extending up until June of 2009, on SES and biomarkers of physiological systems, was conducted by Dowd, et al., (2009) to address SES, cortisol levels and indirect measures of allostatic load.  A total of 26 studies met the inclusion criteria of reported associations between an indicator of SES and cortisol, and/or allostatic load.   Fourteen of these studies utilized salivary cortisol secretions.   The findings overall were mixed with little evidence that lower cortisol related to lower SES and lower SES related to higher allostatic load measurements.   Overall, the authors found more studies with no associations of cortisol to SES than the intuitive finding of lower SES associated with higher cortisol levels.   The unexpected findings were attributed to differences in the nature of this relationship or inconstancies in measurements and analysis of both cortisol and SES.

Standardization of cortisol procedures and analysis, variations in SES indicators used, and the exclusion of subjects using stimulators of cortisol such as smoking, were major recommendations for future research.   Although the review focused on cortisol level daily patterns and indirect measurements of allostatic load, both laboratory stress induced and dexamethasone challenges, were felt to provide more controls for research on the differences of SES on HPA function.

Pluobidis, Benova, Grundy & Layton, 2014, identified four major hypotheses from the literature about the associations between SES and later life health: 

1. Early life SES directly effects later life health.

2. Early life SES indirectly effects later life SES.

3. Early life and later life SES effects health through accumulation of risk.

4. Early life health indirectly effects later life health via later life SES.

A sample of aged 50-53 years was taken from an English longitudinal study on aging.   Multiple measurements of early life and later life SES, health, and fibrinogen levels (indicator of aging) were obtained. The four major hypotheses were compared through statistical modeling. In general, results found early life SES extends directly until the beginning of old age and predicts health at age 65 and older yet fibrinogen levels will vary.

Co-existing with SES are social risk factors defined as (Msall, Sullivan & Park, 2010):  “suboptimal home and community environment, poverty, domestic violence, drug addictions, crime, hunger, and poor quality housing (p 224).”   The conditions of low SES along with access to care issues, coping with multiple adversities, helplessness and low self-esteem; contribute to the risk of preterm births (p 225) along with ethnicity, family history, maternal characteristics, multiple pregnancies and air pollution (Johansson & Cnattigius, 2010; Msall, Sullivan & Park, 2010).

Low SES is consistently associated with poor health and disease yet how this gets translated into biological risk is uncertain and studies have shown inconsistent and at time weak results (Pluobidis, Benova, Grundy & Layton, 2014; Senterfitt, Long, Shih, & Teutsch 2013; Baber, Muzaffer, Khan, Imdad, 2010; Dowd, Simanek & Alello, 2009; Wilkinson & Marmot, 2003; Lupien, King, Meaney, & McEwen, 2001).

Methodological issues in past studies have helped confound the influence of SES and effected some interpretations of DOHaD theory evidence.

In this study, participants’ selection criteria at birth, as designed in the original research, involved representation from all SES groups in each variable of birth status (preterm and full-term born) to control for this effect.  Additionally later SES status at age 23 was assessed for possible individual differences and SES group variations from the prenatal time period.   Multiple measures of SES status, in addition to income, were used and included standardized instruments, education level, occupation level categorization and neighborhood ratings (Farrington, 1991).

Race and Ethnicity

The Center for Disease Control (CDC) found differences in preterm birth by race and ethnicity and statistically analyzed the relevant differences, using percentages, and z tests at the 95% confidence level (Martin & Osterman, 2013).   The percentage results were: 

1. Black infant preterm birth rate (17.1%) was 60% higher than for White infants (10.8%).   American Indian/Alaska Native (13.6%) and Hispanic (11.8%) preterm birth rates were higher than White infants.

2. Black infants had double the early preterm birth rate (6.1%) than Whites, 25% higher than Hispanics (Reagan & Salsberry, 2005), and other ethnicities (2.9%).

3. Black infants were 40% “more likely to be born late preterm than White infants with Hispanic infants more often than White infants, (Martins & Osterman, 2013).”

Reagen and Salsberry (2005) studied the health disparities of preterm births among Blacks, Hispanics, and Whites focusing on social contexts of neighborhood disadvantage and cumulative exposure to income inequality while controlling for individual risk factors.   Neighborhood poverty and housing vacancy rates increased the rates of premature births for Blacks while income inequality directly effected only Hispanics (Reagan & Salsberry, 2005).   Not withstanding these findings, the close relationship of social risk factors from low SES to ethnicity (Msall, Sullivan & Park, 2010) confounds the separation of SES from ethnicity effects.  Likewise, other epidemiological studies have shown, that among the multiple causes for spontaneous preterm births, being a member of the Black race, is also a risk factor (Goldberg, Culhane, Iams & Romero, 2008).

Mustillo, Krieger, Gunderson, et al., 2004, found self-reported experiences of racial discrimination by Black women were related to premature birth weights and may contribute to disparities in perinatal health between races (Black 50%: White 5%). One landmark study accounting for social disparities, showed that even college educated Black women have an increase rate of premature births when compared to White college educated women (Schoendorf, Hogue, Klieinman, et al., 1992). In past racial and ethnic disparities research, studies have separated their focus on either the social construct of race or the biological processes (Kramer & Hogue, 2009).

A systematic review of the research literature, in 2009, focused on integrating the racial biological and social patterning of premature births, with the expressed purpose of  ‘understanding the etiology of black-white racial disparities in preterm birth (Kramer & Hogue, 2009, p 85)”.   Over 1,459 citations were reviewed spanning from 1960-2009. Studies utilizing ultrasound-based measurements and data cleaning  methodology approaches that decreased misclassification were utilized.  Conceptual framework reviews lead to 3 primary biological pathways mediating the racial disparities in preterm birth:  placental dysfunction, HPA dysfunction and maternal-fetalinflammation.   Pre and peri-conceptual maternal health as well as genetic and epigenic pathways studies were included in the review.  Overall the researchers found evidence to support socially patterned maternal stress as a cause of racial disparities.   The identification of few studies addressing genetics and the challenges of controlling for multiple causal explanations prompted the authors’ suggestions for future research.  The suggestions included incorporating biological markers into socially focused preterm birth studies as well as improved epidemiological design studies.

In this secondary analysis longitudinal study, the original racial and ethnic composition of the sampling is predominately White and reflective of the population and geographical location in Southeastern New England from 1985 to 1989 (Sullivan, et al., 2008).  The homogenous composition of this study population will be applicable to the White racial group and not reflect the disparities inherent between Black and Hispanic populations.   The identification of the White racial group preterm health outcomes from 1985-1989 may contribute to further knowledge of health advances made since that time period and could possibly be utilized to compare the associated magnitude of racial disparities today.

Gender Globally, males are slightly more likely to be born prematurely than females (Katz, Lee, Kozuki, et al, 2013).   Decades of past research have shown males born prematurely have higher mortality and morbidity and the phenomena is often referred to as “male disadvantage” (Brothwood, Wolke, Gamsu, Benson & Cooper, 1986; Stevenson, Verter, Fanaroff, et al., 2000; Banga, Barche, Singh, Sheehan & Vasylyeva, 2015).   The confirmed risks of high blood pressure and placenta abnormalities to the pregnant mother carrying a male fetus is thought to occur secondary to sex differentiation hormones in utero and at conception (Katz, Lee, Kozuki, et al., 2013, Ingemarsson, 2003).   Moreover, Sweden national figures show death rates are higher for males by 55- 60% when born between 23 and 32 gestational weeks (Ingemarsson, 2003).   Immediate complications of respiratory distress syndrome are greater for prematurely born males and cognitive recovery after intracranial hemorrhage is less when compared to premature females (Imaemarsson, 2003).   Similarly, in the United States, males are also more likely than females (OR = 1.21; 95% CI: 1.02 – 1.42) to be born at 33 to 36 weeks (McGregor, Leff, Orleans & Baron, 1992). To put it another way, if a male and female are born at the same prematurely gestational age then male infants risk becoming more seriously ill than females (SMFM, 2015).

While prematurity survival rates have increased, the prematurely born at 25 weeks will develop disabilities (1:10) such as lung disease, cerebral palsy, blindness or deafness;  50% disabilities; and more commonly cognitive and neurological impairments (Banga, Barche, Singh, Sheehan & Vasylyeva, 2015).   A retrospective chart review of 160 (male 59% and female 41%) pediatric records at a Texas clinic focused on children and adolescents born prematurely and any gender differences in medical diagnoses.   The sample consisted primarily of White (39.2%) and Hispanic (38.0%) races born prematurely and between 10 and 21 years old at the time of the chart review.   Gestational ages were divided into two groups of 32-37 weeks and < 32 weeks with birth weight divided into 7 groupings ranging from extremely low birth weight to large for gestational age.   The incidence of neonatal complications between genders was assessed according to:  jaundice, metabolic complications, respiratory distress syndrome, sepsis, intracranial hemorrhage and hypertension.

Long-term morbidities included ICD-9 diagnoses of: asthma, allergic rhinitis, cardiac defects, behavioral issues, developmental delays, growth delay and kidney anomaly and diseases.   Even though, more preterm born males were at weights appropriate for their gestational age, the study found males had a higher incidence of neonate complications especially:  jaundice (63.1 vs. 36.8%; p = 0.02), metabolic issues (64.2% vs. 35.7%, p = 0 .03), and respiratory distress syndrome (60.5% vs. 39.4%, p = 0.02),  (Banga, Barche, Singh, et al., 2015).   In contrast, prematurely born females weights were primarily small for gestational age.   No differentiation between genders for neonatal diagnoses of intracranial hemorrhage, sepsis or hypertension were identified.   The only significant gender difference in long-term morbidities found was notably in behavioral issues for males and mostly diagnosed with attention deficit hyperactivity disorder (6% vs. 2%; p < 0.01), (Banga, Barche, Singh, et al., 2015).

The prematurely born female chances of survival are higher than males yet Black prematurely born females, weighing about 2.2 pounds or less, have a higher rate of survival than their White peers (UF, 2006).   Researchers (UF, 2006) who studied vital statistics from Florida, between the years of 1996 and 2000 utilizing records of 5,076 babies born in the state, found females at lower birth rates faired better with Black females fairing better overall.   Although the researchers are yet unable to explain why this racial and gender phenomena exists it is known that female premature babies in general have more developed lungs than males (UF, 2006).

Finally, there are a few decades of history of the vulnerability of preterm males over preterm females for increased mortality and morbidity.   Studies have shown despite advancements in preterm neonate care this phenomenon continues to exist as a risk to full term pregnancy and neonatal complications with male births.   A few endocrine and biological adaptation explanations have been proposed without any known etiology of this gender-related health disparity (Banga, Barche, Singh, et al., 2015). In this study, overall gender differences as well as gender and birth status interaction are analyzed to identify the direction and strength of this effect (Baron & Kenny, 1986).

Prematurity and Development at Age 23 as Emerging Adults

Multiple factors including SES, race, gender, environment and lifestyles, to mention a few, influence the health and the formation of physical diseases and emotional health in all adults. The prematurely born Age 23 emerging adult entered this world with the disadvantage of multiple immature bodily systems.

Additionally, as has been previously stated, human brain growth occurs during uterine development (Phillips, 2007), the newborn’s central nervous system (CNS) evolves rapidly and peak growth spurts continue well into the mid-twenties of age (Epstein, 1986) resulting in multiple sensitive time periods of critical influence. In preterm infants who require extensive neonatal intensive care, it is possible that the HPA Axis is repeatedly activated which may result in permanent programming of early life responses (Maniam, Antoniadis, & Morris, 2014; Reynolds, 2013; Sullivan, Hawes, Winchester & Miller, 2008).  Routine medical procedures activate the newborns stress response system to react and moderate levels of endocrine hormones have been found (Jensen, Beijers, Riksen-Walravan & de Weerth, 2010) that contributes to alterations within the HPA Axis or fetal programing from cell death, and failed or delayed responses of the central nervous system (CNS),  (Sullivan, Hawes, Winchester & Miller, 2008).  See Figure 5

Figure 5 Prematurity, Postnatal Stress and HPA Function

Figure 5. Reprinted by permission Mary. C. Sullivan. 2008-2013. In “Risk and protection in trajectories of preterm infants: Birth to adulthood (Grant # NIH R01 NR003695-14).”  Bethesda, MD, National Institutes of Health, National Institute of Nursing Research.

Structural differences in the brains of premature low birth weight infants with alterations continuing into adulthood have been found by researchers (Nostarti, Murray & Hack, 2012).   Measurements of brain pathology using a variety of biomarkers, such as salivary cortisol levels as an endocrine marker of the HPA Axis activation is in wide use and may serve as a transitional marker for psychological pathology (Turner-Cobb, 2005).

Feldman, Weller, Sirota & Eidelman (2002) in Israel, studied the effects of mother to infant (or Kangaroo Care) skin-to-skin contact on both prematurely and full term born infants.   Specifically they examined “infants’ capacity to regulate sleep and wake states, organize behavior, regulate negative emotions, modulate arousal, coordinate attention to mother and an object, and sustain effortful exploration of the environment (p 194)”.   The infants born prematurely showed an abnormal response to stimuli and often could not inhibit reactions.   Three theoretical perspectives were included:  There is unique time windows for input required for optimal development of the central nervous system and behavioral organization.   Sensory development is sequential and lastly maternal proximity organizes sleep, rest and behavior inhibition.

A prematurely born group with intervention (n = 73) and a matched control group (n  = 73) without intervention was used.   Pre and post interventions as well as multiple time point measurements were used.   The premature infants who received skin-to-skin contact from the mother were found to benefit by improved behavior organization and emotional regulation when they reached full term.   Hence, self regulation and the ability to adjust behaviors to the situation is a challenge to the preterm infant, requires environmental control and sets up regulation parameters overtime (M.C. Sullivan, personal communication, November 11, 2014).

Affect regulation of emotional experiences to serve a purpose or goal contribute to meeting developmental milestones and adult maturation while emotional influences on decision-making have been found during adolescent to contribute to behavioral (alcohol and nicotine dependence), emotional and clinical disorders (Dahl, 2001).  Prematurely born children are at increased risk for behavioral and emotional health problems along with associated psychiatric disorders especially anxiety, depression, inattention and social difficulties (Johnson & Marlow, 2011; Strang-Karlsson, 2011). Children born at extreme prematurity have been found to experience a 50-70% higher rate of attention and behavioral problems in school despite normal IQ scores (Lynn, Cuskelly, O’Callaghan & Gray, 2011). Even children born near term  (34-35 weeks gestation) have a 36% increased risk for developmental delay or disability at kindergarten age (Rabin, 2009). Preterm born 17-year-old late adolescents have a higher percentage of psychological problems when compared to the United States national age-related statistics (NIMH, 2007a & 2007b; ADAA, 2007):  11% with ADHD compared to 3-5%; 12.1 % diagnosed with depression compared to 5%;  9.8% diagnosed with anxiety disorders compared to 3.1%  (Sullivan, 2008).   Thus, prematurity is associated with behavioral and emotional health issues, as well as, psychiatric disorders from preschool through adolescence age, and have risk for continued problems in adulthood.

Miller, Sullivan, Hawes & Marks (2009), reported on their prospective, longitudinal sample of 186 children, at age 12, grouped into four preterm perinatal morbidity groups (healthy preterm without medical or neurological illness, medical preterm with clinical illness but without neurological abnormality, neurological preterm with severe illness and small for gestational age preterm with or without medical problems) and healthy full-term comparison group.   A variety of biological, social and physical environmental factors were measured utilizing a battery of tests and neonatal medical data sources.   Differences for neurological status, motor status and health at age 12 were significant with abnormal high rates in the four preterm groups compared to the full term group.   Total health outcomes of the four preterm groups were 3.4 times more likely than term births to have overall abnormal health status at age 12.

A secondary analysis of this data (Wright & Sullivan, 2011) demonstrated that prematurity measured by birth weight was associated with childhood psychiatric symptoms at age 12.  Additionally, the mother’s perception of their premature child’s vulnerability and psychiatric symptoms correlated positively at ages 4, 8, and 12.

Sullivan, Msall & Miller, (2012), found a higher percentage than the United States’ statistics for psychological problems in their age 17 cohort of the study related to attention deficit hyperactive disorder (ADHD), depression and anxiety.   This prospective study reported on the same cohort at age 17 consisting of 215 infants born between 1985-1989 with preterm birth weights <1850 grams and grouped by neonatal morbidity then compared them with a full term group.   Outcomes of functioning and disability included body functions, body structures, activities and participation.

Contextual factors were identified according to the World Health Organization International Classification of Functioning (The ICF Model, WHO, 2002).   At age 17, 180 of the 215 adolescents, completed the standardized assessment process that analyzed health, neurological, chronic conditions, psychological, environmental-socioeconomic, personal-neonatal morbidity and gender status.   Overall results indicated that physical long-term effects of prematurity were not only confined to infants with very low or extremely low birth weight but also included small for gestational age and preterm infants without neonatal complications.  Higher percentages of psychological problems, 11% ADHD (4.1%), 12.1% (5-8%) depression, and 9.8% (3.1%) anxiety disorders were found.

Emerging Adults The age period between 18-25 years old is now termed the period of “emerging adulthood” due to cultural delays in reaching developmental milestones (Arnett, 2013). In the past, these ages were considered part of young adulthood, when the taking on of adult roles were not delayed. Emerging adulthood, is now a separate category in the life span characterized by:  identity exploration, instability, self-focus, feeling in-between adolescence and adulthood and feeling hopeful about future possibilities (Arnett, 2013).  More importantly, emerging adults do not exist in all cultures and only exist in cultures (middle income) that allow the putting off of adult roles and responsibilities (Arnett, 2013).

Recent national behavioral trends in transitioning into adulthood have shown delays in traditional major life events such as age at marriage and parenthood, instability of residence, higher rates of enrollment in college education and a decline in emerging adults working for pay (Ingels, Glennie & Lauff, 2012; Arnett, 2013).

Arnett (2000 & 2013) characterizes the self-views of emerging adults, in addition to not perceiving themselves as an adolescent or an adult, as also not fully accepting responsibility and not making independent decisions.   This may be especially difficult for those who were born prematurely.    In addition during the transition from adolescence (ages 10-18) into emerging adulthood (ages 18-25) extensive related endocrine system changes occur and influence bodily processes (Arnett, 2013). This turbulence in endocrine hormones, involving all brain structures, adds to the life experiences influencing brain growth and emotional health.   An emerging adult born premature may not have the flexibility or adaptability of the HPA axis responses to achieve allostasis and resiliency.

Globally emerging adults are experiencing life as less meaningful and health professionals are increasingly concerned about emerging adults negative behavioral choices to deal with stress as a way of coping (Hutchinson, Stuart & Pretorius, 2007).

American adolescents and emerging adults have a higher rate of risk behaviors than other countries (Arnett, 2013).  Additionally, emerging adult college students were more likely than older students to become angry or hostile about negative life events instead of becoming more anxious and depressed (Jackson & Finney, 2002).

Emerging adults born premature who may have physical difficulties, learning problems and limitations in social skills have an additional level of coping complexity during this developmental period (Sullivan, 2008).   Considering these challenges at age 23, it may be expected that prematurely born emerging adults will have difficulty coping with adult stressors.

During emerging adulthood, exploration and changes occur that often lead to lasting life choices (Arnett, 2000) with stressors, coping styles and neurophysiological responses of these life choices effecting overall health (Lovallo, 2005; Somerfield & McCrae, 2000). Differences in stress exposure, appraisals of stress and coping styles have been identified in adults with immune system disorders, cardiovascular, depression disorders, and include a variety of physical and mental diseases (Cohen et al, 2007; McEwen, Gray & Nasca, 2015; Segerstrom & Miller, 2004).    Anxious adults with comorbid depression have been found to use more emotion oriented coping than individuals without a comorbid diagnosis (Man, Dugan, & Rector, 2012).

The role of avoidance coping has been associated with the generation of stress that a decade later leads to depressive symptoms (Holahan, Moos, Holahan, Brennan & Schutte, 2005).

Given that higher percentages of psychological problems were found at the age 17 original study time point of both gestationally small and premature infants and that growth of the brain was not fully completed, then obtaining stress, coping, emotional intelligence and emotional health measurements by examining the age 23 cohort will provide additional developmental data.

Stress and Coping Multiple definitions of stress and stressors exist and are often criticized as being “circular, formless or varied (Aldwin, 2009)”.   Indeed, stress is defined according to each multiple discipline perspectives and applied to divergent topics (Aldwin, 2009).

Stress definitions have also been classified as:  a response (Styles:  physiological), a stimulus (Holmes & Rahe:  adaption to life events), a transaction (Lazarus & Folkman:  appraisals and coping), a dynamic process, state-like or trait-like and acute or chronic (Rice, 2012; Lyons 2012; Butler, 1993).

Acute and chronic psychological stress in this study are defined by the DOHaD theory. Acute stress is characterized by the ‘fight or flight” response and is short term with transient over arousal, gastrointestinal symptoms, muscular problems and combinations of the three stress emotions of anger or irritability, anxiety or depression (Miller & Smith, 2015).   The American Psychological Association (2015) also defines an “episodic acute stress” characterized by an “individual always in chaos, in a hurry, full of nervous energy and type A personalities.”   Chronic stress, is the daily wear and tear of allostatic load and can lead to “suicide, violence, heart attack, and stoke (Miller & Smith, 2015)”.

Related to types of stress are the Diagnostic and Statistical Manual of Mental Disorders, 5^th Edition (DSM-5; APA, 2013) broad definition of anxiety disorders.

Anxiety disorders have “features of excessive fears, anxiety and related behavioral disturbances (p 189)”.   “Fear is the emotional response to real or perceived imminent threat, resulting in arousal of the flight or fight response, and anxiety is the anticipation of the future (p 189),(APA, 2013).”

Measuring stressors is often done by checklists and interviews (Gutman & Nemeroff, 2011). A number of functional tests of HPA activity have been developed including the dexamethasone (DST) suppression test, the corticotrophin-releasing factor (CRF) stimulation test and the combined (DEX/CRF) test. These tests are invasively administered or use oral administration (Gutman & Nemeroff, 2011). The Trier Social Stress Test (TSST) a reliable non-invasive psychological challenge test was developed to test HPA axis reactivity to psychological stressors (Gutman & Nemeroff, 2011). A standardized laboratory protocol involves a 10-minute public speech and a mental arithmetic test. The TSST combines both uncontrollability and evaluative threat, which are the conditions of DOHaD theory to activate the HPA axis response (Kudiela, 2008). As a result the response to TSST is both physiological and psychological including increased anxiety and negative mood (Kirschbaum, 2010).

TSST testing has shown that total plasma cortisol levels are overall higher in elderly women when compared to elderly men (mean age 67.3 years), younger men and women (mean age 23.5 years) and without any differences in the patterns of reactivity (Kudielka, Buske-Kirschbaum, Hellhammer & Kirschbaum, 2004).

Schommer, et al., (2003) found habituation to psychosocial stressors unchanged overtime even if an individual was found to be a “high” or “low” cortisol responder to the TSST.   Meaning, that the cortisol response to stress pattern remains uniform when activated with repeated stress. Additionally differences in cortisol response to TSST have been found in many psychological disorders (McGirr, Daiaconu, Berlim, Pruessner, Sable, Cabot & Turecki, 2010).

Maltreated female youths (ages 12-16 year) show a different cortisol pattern in response to the TSST (McMillan, et al., 2009).   Youth without a history of maltreatment showed an increase of cortisol followed by a gradual decline, in contrast to youths with a history of maltreatment, who showed no increase in cortisol in response to the TSST and no decline.   Emerging adults, at age 23, who were born preterm are expected to have a pattern of salivary cortisol response of little to no increase and little to no decline after exposure to the TSST.   Total cortisol levels may vary in comparison to the full term born by being higher.

Lazarus and Folkman’s enduring model (Carver, 2011) of psychological stress and coping responses focuses on the process between the person, environment, their appraisal of and context of the event and is consistent with DOHaD theory (Lazarus, 1999; Park & Folkman, 1997). Psychological stress is what happens when a person is defeated in obtaining their goal commitment, intention or values (Lazarus, 1999).

Stress mediating processes include coping responses defined as problem and emotion focused resulting from the primary and/or secondary appraisals used and are considered psychologically effective if they work for the individual (Lazarus, 1999).

Moos and Holahan, (2003), added dispositional (inclined to cope) factors resulting in two orientations:  focus and method of coping. The focus is either approach or avoidance while the methods of coping are cognitive and behavioral (Billings & Moos, 1981; Moos, 1995; Moos & Holahan, 2003).   Approach coping is defined as problem solving attempts to deal with and resolve stressors while avoidance coping is defined as emotion focused attempts to avoid dealing with, thinking about, and managing emotions associated with the stressor (Moos & Holahan, 2003).

Combining the focus and methods of coping results in eight coping subset styles listed in Table 1 and defined in Table 2, and are measurable with standardized self-report instruments.

Moos and Holahan’s definition of coping is one of three models (Aldwin, 2009; Rice, 2012). Two of the three other models explaining coping are:  the environmental system, addressing ongoing life stressor and social resources, and the personal system involving individual demographics (Rice, 2012).   The third model, transactional, incorporates personal and contextual factors of the stress situation with coping skills that affects the psychosocial functioning and maturation at the individual’s next stage of development (Rice, 2012).   Determinants of coping responses are the types and the severity of the stress, social resources and demographics (Rice, 2012). Gender, age, education, SES, cognitive abilities, problem-solving skills and personal control or regulation will effect both the focus and method of coping.

Type of Coping            Approach Coping                             Avoidance Coping _____________________________________________________________________________________________

Cognitive                     Logical Analysis                              Cognitive Avoidance Positive Reappraisal                         Acceptance/Resignation _____________________________________________________________________________________________

Behavioral                   Seeking Guidance and Support        Seeking Alternative Rewards Problem Solving                               Emotion Discharge ____________________________________________________________________________________________

Table 1. From “Dispositional and Contextual Perspectives on Coping:  Toward an Integrative Framework,” by R. H. Moos and C. J. Holahan, 2003, Journal of Clinical Psychology, 59 (12), p.1391. Copyright 2003 Wiley Periodicals, Inc.  Adapted with permission.

Table 2

Definitions of Coping Styles (Moos & Holahan, 2003)

___________________________________________________________________

COGNITIVE APPROACH COPING

Combination of logical analysis and positive appraisal strategies Logical Analysis                                         Attention to one aspect of the situation Utilizing past experiences Thinking about possible actions and their consequences Positive Appraisal                                       Accepting situation and restructuring it in a favorable way _____________________________________________________________________

BEHAVIORAL APPROACH COPING

Dealing directly with situation through concrete actions Seeking Guidance and Support Seeking guidance and support from others Problem Solving Actions derived from the approach strategy, and organized information about the problem along with the allocation of resources, monitoring progress and evaluating results (Cherry, 2013)

COGNITIVE AVOIDANCE COPING

Combination of cognitive avoidance and acceptance or resignation.

Cognitive Avoidance Minimizing or denying the situation and/or consequences Acceptance/Resignation Accepting the situation and/or consequences and deciding it cannot be altered.

____________________________________________________________________

BEHAVIORAL AVOIDANCE

Involves seeking alternative rewards, venting feelings and/or engaging in risky behaviors.

Seeking Alternative Rewards

Replacing loses with involvement in new activities resulting in an alternative source of satisfaction Emotional Discharge

Openly venting feelings of anger and frustration Behaviors that temporarily reduce tension involving acting impulsively and/or doing something risky _____________________________________________________________________

Table 2. From “Dispositional and Contextual Perspectives on Coping:  Toward an Integrative Framework,” by R. H. Moos and C. J. Holahan, 2003, Journal of Clinical Psychology, 59(12), p.1391. Copyright 2003 Wiley Periodicals, Inc.  Adapted with permission.

Gender Differences in Stress and Coping

The American Psychological Association  (APA, 2010) reports differences in the way males and females react to, manage, and view their ability to deal with stress.   In general:  females report more stress than males (especially married females), more physical symptoms of stress, and identify money and economy more often as sources of stress while males report work more often as a source (APA, 2010).

Matud (2004) examined stress and coping differences between genders and found outcomes consistent with decades of past research.   There were no significant differences in life event frequencies experienced between genders, yet females reported life events as more negative (p = <0.05), less controllable (p =<0.01) and also reported more chronic stress (p = <0.01) than males.   The major stressors for females were family and health related while males’ stressors were focused on relationships, finances and work events.   Females utilized emotional and avoidance coping styles more frequently while males used less emotional coping (p < 0.001).

Females experienced more psychological distress (p < 0.001) and somatic symptoms (p = <0.001) than males.

Tamres, Janicki and Helgeson  (2002) in their meta-analysis of gender coping differences reviewed 50 English language empirical studies between the years of 1990-2000, with actual stressor measurements rather than hypothetical situations, to examine coping responses.   The authors categorized the studies’ definitions of coping behaviors into problem-focused and emotional-focused with avoidance, approach-based and other components.   Females showed a wider repertoire of types of coping behaviors with higher usage of most types than males. Significantly, females used more verbal, social support seeking, ruminating and positive self -talk than males. A few differences between genders were found with females more often than males using avoidance responses for coping with other’s health stressors while males used avoidance responses for coping with relationship stressors.

Gender differences in coping patterns at emerging adulthood (ages 18-25 years) have been found to reflect established cultural qualities (Davis, Burleson & Krusewski, 2011; Kim & Sasaki, 2014) with endorsement of masculinity as a significant predictor in problem-oriented coping (Lipinska-Grobelny, 2011). A recent decline in female gender depressive symptoms in emerging adults is thought to be associated with greater autonomy and empowerment (Frye & Liem, 2011) although in general, females had been found to have higher emotional coping than men (Durm & Glaze, 2002).

DOHaD, Stress and Coping DOHaD theory focuses on the HPA Axis physiological and biochemical mechanisms and how these interact, as well as, the resultant effects on higher nervous system processes such as behaviors, cognitions and emotions. Importantly, DOHaD theory defines stress as a real or perceived threat (physical or psychological), a stressor as an aversive stimuli and the stress response as the activation of numerous complex body systems through the HPA axis.

Figure 6 Relationship of Hippocampus to HPA Axis

Figure 6. Reprinted by permission from Macmillan Publishers, Ltd: Nature Neuroscience, “How Adversity Gets Under the Skin” by Steven E Hyman,2009, Nature Neuroscience, 12, 241-243. Copyright 2009 by Macmillan Publishers, Ltd.

Accordingly, stress, coping responses and affective states are thought to influence physical pathology by direct and indirect influences on the biological processes and behavioral responses with the HPA Axis seen as the gateway (Cohen, Janiki-Deverts & Miller, 2007; Heindel & Vandenberg, 2015; Lovallo, 2005; Somerfield & McCrea, 2000). The roles of allostasis and allostatic load above all, is the link provided as a mediator of the accommodation to the stressor, being that system-wide physiological changes are made and the HPA Axis interconnects with other brain areas including the prefrontal cortex (PFC) to fit or adapt to the situation (McEwen, Nasca & Gray, 2015; McEwen, 1998, Ganzel, Morris & Werthington, 2010), (See Figure 6).   Figure 5 illustrates  “the HPA axis under the excitatory control of the amygdala and inhibitory control of the hippocampus (Hyman, 2009)”.

Importantly, studies of individuals experienced in meditation, show related inverse changes in functional brain imagery and mapping of brain activity (Hoffman, et al., 1982).   When relaxation increases, heart rate, blood pressure and hippocampus signal activity decreases (Hoffman et al.,1982).   Individuals with years of meditation experience also show thickening of the frontal cortex of the brain (Lazar, et al., 2005).

The researchers think genomic charges are taking place from meditation relaxation since, in addition to sympathetic nervous system responses, chemical changes related to gene changes had been identified.   The gene changes are believed to be a counterpart of the stress reduction response (Dusek, et al., 2008).

The resultant system-wide and HPA-Axis responses to stress (appraisal of threat) and the stressor (aversive stimuli) allows for the inclusion of appraisal, cognition and emotional states as key elements of the stress process. Hence, DOHaD theory advances an integrated model of stress through multilevel biomedical and psychosocial models of stress (Ganzel, Morris, & Wethington, 2011).   Inherent in dealing with stress are the efforts to manage the potential and threatening aspects of the situation or circumstances, whether physical or psychological, and any detrimental consequences to health (Moradi, Pishva, Ehsan, Hadadi, & Pouladi, 2011).

Psychologically, coping is defined as changing cognitively and behaviorally to mange internal or external demands that extend beyond the persons existing resources (Lazarus & Folkman, 1984).

Emotional Intelligence A variety of emotional intelligence definitions were found in a systematic literature review, using the keywords  “emotional intelligence”, from the Cumulative Index of Nursing and Allied Health Literature (CINAHL), Cochrane Library, ProQuest Dissertations, PubMed and the Web of Science.   The majority of publications were from the disciplines of psychology and business with proliferation of the topic in psychology beginning in the early 2000’s (Matthews, Zeider & Roberts, 2004). Recently, the nursing literature on EI has shown an interest in the epistemology, application to educational and the professional competences uses of the construct (Por, Barriball, Fitzpatrick & Roberts, 2011; Akerjordet & Severinsson, 2010; Smith, Profette-McGrath & Cummings, 2009; Freshwater & Stickley, 2004; McQueen, 2004). A large body of literature exists in the business arena where EI is extensively used in hiring, employee assessments, promotions and education of managers (Ashkanasy, Ashton-James, & Jordan, 2004; Freedman, 2010).

One aspect of EI in which all agree is the involvement of an emotional awareness of self and others and emotional regulation or management (Armstrong, Galligan, & Critchly, 2011; Matthews, Zeider & Roberts, 2007 & 2004; Bar-On & Parker, 2000).

A broad common definition of EI is the capacity to identify, process and manage emotions to contribute to a successful life (Armstrong, Galligan, & Critchly, 2011).

Emotional regulation is used interchangeably with emotional management and is defined as the conscious or unconscious control of emotion, mood, and affect that is most commonly achieved through coping strategies (Kowalczyk, 2015; Thompson, 1994). The construct of EI developed out of the research on intelligence to bridge a gap in the role of emotions and is widely used today in education and business settings (Mayer, Salovey, & Caruso, 2004).

Multiple studies have indicated a relationship between stress, coping styles, EI and emotional health (Por, Barriball, Fitzpatrick & Roberts, 2011; Tricky, Far hall, Wertheim, Hinch & Ong, 2011; Ciarrochi, et al., 2002; Gerits, Derkson, Verbruggen & Katzo, 2005).   Studies have found that individuals with high EI report less perceived stress, have better health and report feelings of well being (Ciarrochi, et al., 2002; Hertel, Schutz & Lammers, 2009; Pau and Croucher, 2003). Individuals with low EI reported more stress and difficulties with coping (Gohm 2005; Schutte, Malouff, Thorsteinsson, Bhullar, & Rooke, 2007). Cirrochi, Deane & Anderson, 2000, found stress associated with higher reports of depression, hopelessness, and suicide ideation in college students (N = 302) with high emotional perception scores and found more suicide ideation in college students who scored low in managing others emotions.

The reciprocal relationships between stress, coping and the hippocampus and how it relates to emotional intelligence warrants further investigation as recent studies are finding smaller hippocampal volumes in individuals with long-term Post Traumatic Stress Disorders (PTSD), (Woodward, et l., 2006), (See Figure 6).   Impairment has been found in fear responses of individuals with PTSD involving dysfunctional activation of the HPA Axis and especially within the ventromedial prefrontal cortex (Milad et al., 2009).   EI can predict health functioning as well as distress and traumatic stress (Singh & Sharma, 2012).   Additionally, oxidation of cells occurs from psychosocial stress and reflects an increase in cell aging (Fricchione, 2015).

Singh and Sharma (2012) also found high levels of salivary cortisol improve coping strategies in individuals with high EI while Tang, et al., (2007) found a significant decrease in stress-related cortisol with 5 days of 20-minute meditation improving attention and self-regulation.

Infants born prematurely have a variety of health impairments and a prevalence of neurodevelopmental disabilities in general (Allen, Cristoalo & Kim, 2011). Attention issues and ineffective executive functioning has been found in premature low birth weight infants suggesting that regulation of these functions in the cerebral cortex may have been impaired (Kessenich, 2003).   This area of the brain is also utilized in regulating emotions, interpreting and organizing information for responses which influences the level of an individuals’ EI (Contrada & Baum, 2011). Premature emerging adults who may have perinatal injury and experienced sensitive time periods of stress effecting the HPA Axis, then are exposed to increasing vulnerabilities to stress, coping, and allostatic load during this time period, may have responses that negatively effect EI (Allen, Cristofalo & Kim 2011; Hack, 2009; Ozer & Benet-Martinez, 2006).   EI studies of emerging adults, ages 18-25 years old, are scarce and the majority of studies are cross-sectional with mixed adult age ranges of the participants.   Current measurement of the abilities to lead a successful life often are functional and relate to developmental milestones determined by age-appropriate roles, succeeding in schoolwork, succeeding in the work environment, social maneuverability and forming intimate relationships to mention a few (Sullivan, Msall, & Miller, 2012). Given the identified vulnerabilities of prematurely born infants then measuring and describing EI in this population at age 23, including gender differences, may lead to further understanding of possible predictive functions, critical time periods of risk, protective factors and types of effective clinical interventions. “What is not known is what are the long-term outcomes for adults who were born prematurely and how does stress, biology, caregiving and social factors over a lifetime effect the developing premature born infant? (Sullivan, 2008-2013)”.

EI Theoretical Perspectives Historically, the modern origins of EI can be traced back to John Dewey (1902) “moral motive” and then Thorndike’s 1902 conceptualization of “social intelligence”, which at a rudimentary level, involves the ability to understand and mange other people while getting along socially (Cantor, 2000; Zeidner, Matthews & Roberts, 2009).  Social intelligence was later eclipsed by interest in EI that included differentiating between cognitive, emotional and other factors. EI began to capture scientific interest in academic journals during the early 1990’s through publications by Mayer and Salovey’s model as a way of joining emotions to intelligence studies (Mayer, Salovey & Caruso, 2000).   Culturally, Goleman popularized EI in the 1995 publication of the book Emotional Intelligence that integrated emotions with intelligence.  The importance of both emotion and reason intuitively appealed to the public and some in the scientific community (Matthews, Zeider and Roberts, 2004).

Although the scientific community is studying EI there is not agreement on whether it is a scientific endeavor or if there is a science of EI (Zeidner, Matthews & Roberts, 2009).  A discussion of the theories, measurements, and applications of EI to this controversy follow.

The term EI is used in three ways (Mayer, Salvey, & Carusi, 2000): 

1. The popular meaning is that reason and emotion can be personally integrated and an individual can achieve self-improvement.

2. To describe personality traits through connecting parts of the mind to life outcomes from mental mechanisms, models of self, and self-relevant and general traits.

3. To define a set of abilities dealing with processing emotional information.  

Three theorists have influenced the development of EI knowledge, the use in practice, and research across multiple disciplines: Daniel Goleman, Mayer & Salovey, and Raven Bar-On  (Smith, Profetto-McGrath & Cummings, 2009).

Goleman’s theories developed from the business sector and popularized the concept of EI through use of work competency measures that covered a broad range of personality measures (Mayer, Salovey & Caruso, 2004).   EI to Goleman consist of five parts: “ knowing emotions, managing emotions, motivating oneself, recognizing emotions in others, and handling relationships, (Goldman, Boyatzis and Hay Group, 2005).”   The five parts have been divided into 25 different competencies and include mapping of these competencies.   Goleman’s model is a combination of theoretical concepts and construct derived from multiple psychological perspectives resulting in a popular, trait and ability mixed model.  The model has wide spread popular appeal and is viewed as over inclusive by some in the scientific community.  Goleman, Boyatzis and Hay Group (2005) developed the self-report, Emotional and Social Competency Inventory (ESCI), to assess the emotional and social competencies of outstanding leaders and provided guideline for using the ESCI.  Courses and certifications in EI are also available. Extensive construct and validity studies within business organizations worldwide have been conducted.  Overall, reliability has shown high internal consistency with limited evidence for test-retest reliability. EI has good construct validity with a variety of personality constructs (Wolf, 2005).

Mayer, Salovey & Carusso, 2000, ability model views EI as operating across both cognition and emotional systems as a “unitary intelligence “ (Mayer, Salovey & Carusso, 2000; DeFabio & Saklofske, 2014). Emphasis is on EI as a “concept of an intelligence that processes and benefits from emotion” and is composed of “mental abilities, skills or capacities”.  The unifying system has 4 branches:  emotional perception (perceive, attend, decipher and express), emotional integration (uses cognitive system), emotional understanding (understanding and reasoning) and emotional management (flexible guidelines to mange emotions).  EI is conceptualized as a mental ability and measured with objective tasks.   The Mayer-Salovey-Caruso Emotional Intelligence Test (MSCIET) measures EI by having the individual perform tasks and solve emotional problems in 4 areas:  perceiving emotions, facilitating thought, understanding emotions and managing emotions (Mayer, Salovery & Caruso, 2004^b). The original measurement had criticisms relating to scoring and reliability and a second version addressing some of these issues are now in use (Mayer, Salovey, Caruso & Sitrenious, 2001).

Bar-On, 1997, generally classified as a trait model and by some others as a mixed model defines EI as an “array of non-cognitive capacities, competences and skills that influence one’s ability to succeed in coping with environmental demands and pressures” (Bar-On, 1997; DiFabio & Saklofse,2014; Por, Barriball, Fitzpatrick & Roberts, 2011).  Bar-On and Parker (2000) view the key elements of EI as stress management and the ability to adapt. The self-report trait EI, in this study, was assessed with the Bar-On EQ-I.   The 133 likert-like item questionnaire provides a total EI score (M = 100, SD = 15), and five composite scores for the principle dimensions with 15 related content scale scores (Bar-On, 2002 & 2006; Di Fabio & Saklofske, 2014; De Weerdt & Rossi, 2012).   The five composite score areas are: intrapersonal, interpersonal, adaptability, stress management and happiness (Bar-On, 2002 & 2006; Di Fabio & Saklofske, 2014; De Weerdt & Rossi, 2012).  The 15 content scale scores are:  emotional self-awareness, assertiveness, self-regard, self-actualization, independence, empathy, interpersonal relationship, social responsibility, problem solving, reality testing, flexibility, stress tolerance, impulse control, happiness and optimism (Bar-On, 2002 & 2006; Di Fabio & Saklofske, 2014; De Weerdt & Rossi, 2012).   The model evolved from personality psychology and is measured through a standardized self-report test that assessesself-perceptions rather than actual abilities (Mayer, Salovey, & Carusso, 2000).   Studies show the EQi has a significantly stronger association with mental health than other measures  (Gohm 2005; Schutte, Malouff, Thorsteinsson, Bhullar, & Rooke, 2007).   Raven Bar-On EQi test is used in this study to assess EI, stress management, adaptability and general mood.

A major issue and controversy with EI definitions is they vary so determining what processes, behaviors and outcomes are related makes operationalization of the construct difficult if not possible and limits hypothesis testing (Bar-On & Parker, 2000).  A construct is deliberately and consciously invented or adapted for a specific purpose and identifies what meets inclusion or exclusion by categorization (Mishra, 2013; WSU, 2013).  In addition, constructs serve as an indirect link between the abstraction defined and observed manifestations, in contrast to concepts which are directly linked with observation (WSU, 2013).   Construct validity includes assessing the test content, response processes, internal structure, association with other variables and consequences of its use (Furr & Bacharach, 2014).  Construct validity in EI is, at best questionable, as definitions of EI measurements are based on multiple theories and not on an accurate and agreed upon representation of the entire domain (Stough, Saklofske, & Parker, 2009).  As with many constructs in the social sciences, face validity and content validity refers to the scores derived from each test item and not the test, and is often also an issue with other social science constructs (Stough, Saklofske, & Parker, 2009).  Conceptualization of EI on face validity is appealing yet construct validity across measurements has not been achieved.  Research has shown that ability based and self-reported trait EI tests are not correlated yet differing models of trait self-reported EI were found to be correlated despite describing similar and not identical constructs (DiFaboe & Saklofske, 2015). EI developed from personality theories and maybe redundant with personality constructs (Stough, Saklofske & Parker, 2009).

DiFabeo & Saklofske, 2014, administered a battery of eight tests to assess emotional intelligence (including Raven Bar-On), self-evaluating, resilience, and life satisfaction, to 164 Italian high school students, and utilized multiple regression for the data analysis. Significant results were found for the ‘Big 5” personality factors followed by trait self-reported emotional intelligence tests as the most predictive of self-evaluation, resilience and life satisfaction.   An equally significant finding, was the higher the self-reported trait EI score then the more positive, in control, able to deal with adversity and emotional resources available were perceived by the individual.

Summary

In summary, the construct of EI, is primarily based on personality theories, and measurements lack construct validity across tests.   On the other hand, EI theories are used in business, educational and health settings, and are capturing components or factors related to success in dealing with stress, coping, emotional management and getting along with others socially.   The Bar-On EQi, captures emotional and psychological functioning, fits with DOHaD theory, and was utilized in this study to assess the emotional development of emerging adults and any differences between the prematurely and full term born.   Although definitions of EI vary and a unified construct validity has not been achieved, EI is in wide use, and the appeal may be the ability to capture assessments of traits of a mature personality as demonstrated in the study above conducted by DiFabeo & Saklofske (2014).   Additionally, emotional intelligence scores level off and usually without further gain in adulthood, and this could be related to personality development theories (Garner, Qualter, & Whitely, 2011).  Thus, the measurement of EI at early or emerging adulthood is very relevant.

The Daily Hassles Questionnaire will provide a standardized measure of usual life stress events.  Moos & Billings (1981) determinates of coping styles provides the coping types and strategies used by an individual in stressful events, and carried over from any previous stressful experiences while fitting conceptually with EI (Moradi, Pishva, Ehsan, et al., 2011. Bar-On and Parker (2000) view the key elements of EI as stress management and the ability to adapt.   Elements from each of these perspectives, a long with coping strategies and stress responses have been implicated in health issues.

Emotional Health Health, well-being, mental health and quality of life are similar concepts, which include biology and the individual’s psychological state, to utilize goal directed behaviors and involve multiple causal networks (CDC, 2013).   Emotional health is defined as an individual   “adjusting to new situations and achieving the desired outcomes”   (GUCCHD, 2016).   Higher emotional intelligence maybe related to better emotional health and is linked to aspects of better psychosocial functioning (Schutte, Malouff, Thorsteinnsson, Bhuller & Rooke, 2007).   Drug use, being bullied and violence are a few of the behavioral risk factors that threaten emerging adults emotional health globally (Lister, Merrill, Vance, West, Hall & Crookston, 2015; Orleans, 2008).   There is a range of emotional behavior dependent on an individual’s adaptation to stress and coping style, resulting in effects on feelings of well-being, functioning and health behaviors (Glanz & Schwartz, 2008).

Childhood physical and emotional abuse is related to childhood psychiatric disorders and also effects cortisol (HPA) stress reactivity (MacMillan, Tanka, Duku, Vaillancourt & Boyle, 2013; MacMillan, et al., 2009).   Children and adolescents mental health, according to the CDC (Perou, et al., 2013) is characterized by meeting developmental and emotional milestones, social development, effective coping skills, and being able to function at home, school and other areas of their life.   A similar definition of emotional health exists for emerging adults (Arnett, 2013).   A number of studies of emerging adults show their identified markers of transition are:  “accepting responsibilities for oneself, making independent decisions and becoming financially independent  (Arnett, 2013; Nelson, 2003).”   Additionally, social and emotional loneliness is common and often used for self-reflection and mood management (Arnett, 2013).   High EI scores were associated social functioning, mental health, and vitality (Rey & Extremera, 2013).  EI has been found to have a moderate positive relationship with coping styles and buffered stress in work situations (Gujral, 2013).

Health experts have agreed that most physical problems begin in the early teens through the early 20’s and relate this to behavior patterns developed during this time period (Arnett, 2013). As a result, these patterns of behavior, become established and effect adulthood responses. Health educational programs, directed at risky behaviors have shown mixed results (Arnett, 2013).   Effective psychosocial protective factors have been identified, such as exercise that improves mood, weakens the stress response and is thought to promote neurogenesis, while reinterpreting negative stress stimuli through cognitive positive reframing involves memory, control of emotion, and underlying neurobiology (Feder, Nestler & Charney, 2009).

Tsaousis & Nikolaou, 2005, through two studies (N = 365 & N = 212) with mean age ranges of 25-36 years old measured:

1. EI and general health in one group and

2.  EI, general health and health related behaviors (smoking, drinking, and exercise) in another group.

EI was found to be negatively associated with poor general health, negatively correlated with smoking and drinking and positively correlated with exercise.

Burnett, et al., in 2011, conducted a meta-analysis of peer-reviewed research article findings from 1995-2010.   The purpose was to look at “the prevalence of psychiatric diagnoses in preterm and full-term children, adolescent and young adults. (Burnett, 2011).”   Of 719 articles only 5 passed inclusion and exclusion criteria resulting in 734 preterm and 634 full term (controls) individuals.   The study reported high odds ratio of risk for anxiety, depression and other psychiatric disorders in the prematurely born.

England’s young adults (between the ages of 18 and 25 years old) were found to have resilience-related coping skills by utilizing social support when compared to older adults (ages > 64 years old) who were more resilient and utilized emotional regulation and problem solving (Gooding, Hurst, Johnson & Tarrier, 2011). EI research has shown an association with better health and is a plausible health predictor (Martins, Ramalho & Morin, 2010;  Schutte, et al., 2007).

Adolescents with visible or invisible chronic health conditions (disabilities) were found to do less well in several psychological study outcomes than adolescents without disabilities (Wolman, Resnick, Harris, & Blum, 1994).   One similar invisible condition is Alexithymia, broadly defined “as an inability to express one’s feelings (Merriam-Webster Medical Dictionary, 2015)”. Some researcher postulate this condition is due to deficits in cognitive processing of emotions resulting in undifferentiated poorly regulated emotions and representative of the opposite of emotional intelligence (Velasco, Fernandez, Paez & Campos, 2006; Taylor, Bagby & Parker, 1991). Recent research developments have associated alexithymia with maladaptive emotional regulation, low EI, reduced rapid eye movements, somatic illness, and disease development (Taylor, 2000; Taylor, Bagby & Parker, 1991).

Given the potential for disabilities due to HPA axis, emotional dysregulation in emerging adults born prematurely, coupled with previously mentioned higher incidences of psychiatric disorders, then a range of developmental functioning, well-being and psychopathologies are possible. Indeed, if self-emotional regulation is a part of EI, then the prematurely born are at risk for self-regulation and mental health disorders especially Attention Deficit Hyperactive Disorder (ADHD), (Sullivan, personal communication, November 11, 2014).

Summary

In this study measures for assessing stress responses, coping types, emotional intelligence and emotional health are utilized.   The TSST provided physiological responses by inducing a moderate stress situation along with salivary cortisol levels to assess HPA regulation.   The DHS provided self-reports of everyday stress experiences as representative of developmental milestones as well as stress magnitude scores in all perinatal birth groups.   The Bar-On EQi (1997), total EQi score indicates the emotional intelligence levels in the perinatal birth groups. The ASR provides a emotional health score for all participants. Comparisons of emotional health and stress reactivity of salivary cortisol levels determine if any differences between perinatal birth groups exist.

HPA Axis and the Stress Response The stress response consists of activation of two brain pathways triggered by the hypothalamus.  The first pathway involves the autonomous nervous system and release of epinephrine causing the physiological response of fight or flight.   This pathway’s response is described by Selye’s theory of adaptation to stress definition of immediate responses (Arnold & Kverno, 2009).

The second pathway involved in the stress response is the anterior HPA Axis.

Stress, directly and indirectly affects the HPA Axis, initiating the stress hormonal response at the hippocampus level (Figure 7). The primary hormone regulating this axis is the glucocorticoid, cortisol (Edmunds & Mayhew, 2009).

The HPA Axis, consisting of the hypothalamus, pituitary and adrenal gland, regulates cortisol secretion from the adrenal glands whenever we are stressed and produces a physiological response (Bruyere, 2009).   Corticotrophin-releasing hormone (CRH) from the hypothalamus acts as a chemical messenger to the pituitary gland.   In turn the pituitary gland responds to the CRH and releases adrenocorticotropic hormone (ACTH) into the blood stream.  The circulating ACTH in the bloodstream reaches the adrenal gland that then produces and releases cortisol. The cortisol circulates in the bloodstream until the demand for it is met.

Physiologically, there is a cascade of effects including increased cardiac output, pupil dilation, and shunting of blood from the digestive tract and kidneys to vital organs ultimately resulting in decreased fluid loss, increased glucose, and decreased brain nor-epinephrine.   Once the demand is satisfied the cortisol signals the pituitary to stop producing ACTH.

Figure 7 Corticotropin Releasing Factor System

Figure 7. Reprinted with permission by Sage College. In W. Lovallo, (2005), Stress and health: Biological and psychological interactions (p 116). USA: Sage Publications. Copyright 2005 by Sage College.

The production of cortisol in the HPA axis is regulated through a negative feedback loop system:

1. Low serum cortisol levels activate the pituitary gland to release adrenocorticotropic hormone or acetylcholine (ACTH).
2. ACTH stimulates the adrenal cortex to increase cortisol.
3. A high level of serum cortisol decreases ACTH production and results in a decrease of cortisol production (Edmunds & Mayhew, 2009).

When the HPA Axis is functioning effectively the endocrine response will be quick and adaptive.   If the HPA Axis is exposed to abnormal allostatic loads, defined as repeated stress, lack of an adaptive response, or an inadequate response then dysregulation occurs or allostasis is not achieved, resulting in variations in the cortisol levels (Barker, 1990 & 2007).

Cortisol is the end product of the HPA Axis chemical transactions (de Weerth, Zijl & Buitelaar, 2003).   Cortisol levels follow a circadian or diurnal pattern in normal adults totaling a secretion level of 10 mg. a day (Edmund & Mayhew, 2009).   Cortisol levels are highest in the early morning, peak 60 to 90 minutes after awakening, and are lowest from evening to midnight (Edmund & Mayhew, 2009).

The processes involved in the HPA Axis negative feedback loop are complex and influenced by a number of multifaceted variables.   The timing of sampling during this process of cortisol circulation will affect the interpretation of results nonetheless standardized timeline comparisons and interpretations are available for adults and children (de Weerth, Zijl & Buitelaar, 2003).

The cortisol feedback loop, also involves circulation to the frontal cortex and limbic areas, which effects affective responses and past experiences (Lovallo, 2005).

Differences in the frontal-limbic system as part of the central cortisol feedback system will in turn influence the differences in the stress response (Lovallo, 2005).

Summary

In summary, cortisol is a key stress response glucocorticoid within the HPA Axis and prefrontal cortex negative feedback loop response.   Cortisol can be accurately measured in the saliva as well as the blood, and patterns of responses are standardized.

Changes in HPA Axis responses may help identify periods of transition into disease states.   Salivary cortisol, in this study, was measured to obtain the HPA stress response pattern in a sample of former premature infants and a term born comparison group at age 23 years.   The cortisol response was examined in association with self-reported stress and coping patterns, as well as emotional intelligence and emotional health.

HPA Axis Function in Prematurity Prematurity rates in the United States contribute to national health care issues and costs both in the short and long term.   The rate of premature birth for two and one-half decades, from 1980-2006, increased by 20% (Martin, Osterman & Sutton, 2010). Twelve percent of all births in 2010, were premature, totaling 523,033 births for the year (CDC, 2010).   Currently the mortality rates for premature newborns with low birth weights has decreased by around 2% due to newer preventative and treatment approaches, but long-term morbidity remains (Child Trends Databank, 2011).   This decrease in mortality rates has increased the rates of immediate preterm birth consequences such as:  adrenal insufficiency, intraventricular hemorrhage, patent ductus arteriosus, respiratory distress syndrome, bronchopulmonary dysplasia, necrotizing enterocolitis and retinopathy of prematurity (March of Dimes, 2008).

The long-term outcomes for these infants of decades of premature birth will continue to influence social, educational, community, and family systems well into the future. Prematurity in general has multiple immediate birth-related health issues, and requires continued vigilance throughout childhood (Saigal & Doyle, 2008; Sullivan, et al., 2008; Hack, Taylor, Drotar, Schluchter, Carter, Andreias, et al., 2005; Saigal, Burrows, Stoskopf, Rosenbaum, & Streiner, 2000).  More recently, prematurity effects are proposed to contribute to the development of adulthood illnesses (Hack, 2009; Saigal & Doyle, 2008).

In addition to the specific issues surrounding prematurity, the study of cortisol levels as an indicator of the HPA Axis functioning in those born at term has implications for application to health issues.   Identifying any difference between premature and full term birth HPA Axis functioning by screening for alterations in cortisol levels in combination with stress, coping, emotional intelligence and emotional health measurements may identify vulnerable individuals of all birth group types before expression of any variations (Dedovic, Duchesne, Dager & Pruessner, 2010).

Cortisol levels in newborns have been hard to standardize due the rapid shifting of development, instability of the newborn system and brain volume growth of 15mL/week between 29 and 41 week’s gestation (Main, 2010).   No normal range of preterm cortisol levels exist because of a variety of fluctuation in preterm and full term physiological homeostasis (Ng, 2011).   There is agreement that around 1 year of age preterm cortisol levels begin to stabilize into a pattern approximating the adult diurnal cycle (de Weerth, Zijl & Buitelaar, 2003).   Under normal conditions, with full term infants, cortisol levels reveal a circadian rhythm with the adult pattern of diurnal decline from morning to evening at around 3 months of age (Turner-Cobb, 2005).

Attempts to collect salivary cortisol levels from newborns have met with issues of lack of sufficient saliva quantity and concerns about interference of administered sweet solutions (Morelius, Nelson & Theodorsson, 2004).   Morelius, Nelson & Theodorson (2004), successfully obtained 113 preterm and term infant’s saliva using less than a 10mL sample.   In combination with altering the detection limit of the radio-immunoassay they were able to analyze very low concentrations of cortisol in newborns without interference from oral glucose solutions.

Jansen, Beijers, Riksen & de Weerth (2010), analyzed 48 peer reviewed empirical studies (1978-2008) providing pre and post stressor cortisol levels at various ages.

In the first 13 weeks or 3 months post-natal the mean cortisol reactivity effect to painful stimuli was the highest and continued to decline with advancing infant age groups (3 & 6 months, 6 & 12 months and 12 & 24 months).

Tollenaar, et al. (2010), obtained daily eye swab cortisol levels from 300 infants at the ages of 6 weeks, 5 months and 12 months. The infant’s cortisol levels declined over the year however intra-individual variability was large and stabilized between 5 to 8 months.

Knowing that the newborn’s system, whether preterm or full term, is continuing to evolve at a rapid pace, understanding the processing of a variety of stressors during critical growth periods become paramount.   Prematurity is a stressor itself due to the immaturity of the central nervous system and the neonatal course, which may include neonatal illness and a long intensive care stay.   Routine medical procedures activate the newborns stress response system to increase cortisol secretion.   Physical stressors to the infants HPA Axis have resulted in subsequent moderate increases of cortisol levels (Jansen, Beijers, Risken-Walraven & de Weerth, 2010). The combination of increased cortisol levels, contributing to cell death and a failed or delayed response of the central nervous system (CNS) results in alterations within the HPA Axis or fetal programming (Sullivan, Hawes, Winchester & Miller, 2008).

Sullivan, Hawes, Winchester and Miller (2008), define the immediate effects of premature birth or neonatal programming and include individual differences such as genetic, developmental and social/care giving, which influence the neonates behavioral and neuroendocrine response (See Figure 8). This response is heightened or hypersensitive and contributes to the neonatal allostatic load.

Figure 8 Premature Birth:  Fetal/Neonatal Programming

Figure 8. Reprinted by permission Mary. C. Sullivan. 2008-2013. In “Risk and protection in trajectories of preterm infants: Birth to adulthood (Grant # NIH R01 NR003695-14).”  Bethesda, MD, National Institutes of Health, National Institute of Nursing Research.

Long-term exposure to repeated negative developmental stress increases the allostatic load on the HPA Axis and if not counteracted by positive or protect factors, such as mediated by mother-child attachment security, then adult outcomes of disease occur. Sullivan, et al. (2008), attributes adulthood disease development to factors both positive and negative, defined as cumulative risk and protection, which affect long-term outcomes of functional, emotional, executive brain and work performance (See Figure 9).

In summary, early trauma with long term neurological consequences that are dysfunctional, especially when paired with the proper timing of a stressor, at a critical time in the development of a child, may result in health, emotional and behavioral consequences exhibited in adulthood.

Figure 9 Cumulative Effects of Caregiving and Social Environments Over Time

Figure 9. Reprinted by permission Mary. C. Sullivan. 2008-2013. In “Risk and protection in trajectories of preterm infants: Birth to adulthood (Grant # NIH R01 NR003695-14).”  Bethesda, MD, National Institutes of Health, National Institute of Nursing Research.

HPA Axis Function in Stress and Anxiety Anxiety is a complex response to a stressor that includes emotional, cognitive, behavioral and physiological responses.   Diagnosed anxiety disorders, which last longer than 6 months, affect 40 million adults age 18 and older each year (NIMH, 2009). Anxiety disorders commonly occur with mental and physical illnesses, as well as alcohol and substance abuse.   Individuals may complain of uneasiness, fearfulness or nervousness, increased heart rate and increased respirations, to mention a few signs and symptoms.

Walter Cannon’s well-known fight or flight response, further developed later by Hans Selye, prepares the body to respond by increasing blood pressure, heart rate, cardiac output and other responses as indicated in HPA functioning noted earlier (Halter & Vascarolis, 2010).   If the individual’s underlying neurological functioning or HPA Axis functioning is dysfunctional then mediating effects, positive and negative or cumulative protective and risks factors, will effect this response according to the DOHaD theory. These mediating stress effects can be from a physical, psychological, emotional, cognitive, intellectual, major life events, environmental and social/caring interactions.

If the individual’s anxiety is transient and without dysfunction then no permanent change will occur in the HPA Axis functioning. If there is an early trauma, such as prematurity, with long term neurological consequences that are dysfunctional, paired with the proper timing of the stressor during critical developmental periods of the child, then the consequences will be evident in adulthood (Turner-Cobb, 2005).   In other words, the adaptability to a stressor or anxiety rather than the initial reaction will predict long-term outcomes.

What is known during childhood, is the relationship of low birth weight in preterm births and subsequent psychopathology is coactive with psychiatric problems but not necessarily medical problems (Nomura & Chemtob, 2007).   Additionally directionality of cortisol levels in childhood is known and has been determined by their behavior of engaging or withdrawing responses (Turner-Cobb, 2005).   Often a child born prematurely will have a blunted cortisol response.   A blunted cortisol response, is defined as a failed, less intense, less concentrated or delayed cortisol response (Bruehl, Wolf, & Covit, 2009; Sullivan, et al, 2008).

Various adult studies exist that measure cortisol levels in a variety of psychological pathologies, emergency situations, and anxiety reduction interventions (Harvey, Nathens, Bandiera & LeBlance, 2010; Rapaport, Schettler & Bresse, 2010; Yu, 2010). At the other end of the life spectrum, Lanze, Mantella, Shi, et al., (2010), report that researchers have found that one-third to one-half of elders develop generalized anxiety disorder later in life, when aging neurological processes interfere with brain cortex messages to the HPA Axis.

Although they are careful to report that multiple neurological pathways affect health, the utilization of the selective serotonin re-uptake inhibitor drug escitalopram was effective in symptom reduction and reducing cortisol levels in subjects that had diurnal pretreatment cortisol levels above normal peak and daily total levels (Lanze, Mantella, Shi, et al., 2010).

Elevation of cortisol levels in healthy adults has been found when the subject is exposed to a standardized transient psychological stress test, the Trier Social Stress Test (TSST) to trigger an acute anxiety or stress response. If a stressor or anxiety response is elicited then the corresponding standardized timed cortisol level responses reflect this change. A baseline diurnal cortisol level is obtained for comparison to the stress induced cortisol response to measure reactivity of the HPA Axis. This psychological stress induced measurement is useful to look at the level of stress or anxiety triggered (and the HPA Axis response) and the intensity level of the cortisol release.

Stress, coping, emotional intelligence and emotional health differences have been found in some studies with prematurely born children.   Given that the prefrontal cortex and limbic system is part of the corticotropin releasing factor system, then involvement of these structures and function effects will also influence emotional, affective, appraisal and coping in the emerging adult (See Figure 10).

Summary

Structural differences in the brains of premature, low birth weight infants with alterations continuing into adulthood have been found by researchers (Nostarti, Murray & Hack, 2010). The concept of allostatic load and research studies related to DOHaD theory links (Sullivan, 2008-2013):

“1. Repeated infant stress responses to prematurity and neonatal experiences with increased allostatic load.  2. The cumulative effects of the social environments overtime with disease formation”.  

Measurements of brain pathology using a variety of biomarkers, such as magnetic resonance imaging and free saliva cortisol levels, have yielded general structural and functional data.

Figure 10 Stress, Coping, Prefrontal and Limbic System, HPA Axis, Cortisol, Emotional Intelligence and Emotional Health

EMOTIONAL STRESSOR

|

APPRAISALS (Aversive)

|

_____________________________________________________________________                                                                      

 

                                 CORTISOL ___   ___   ___   ___   ___   ___   ___   ___   ___

                                               |                                         = Prefrontal Cortex &         |

   Limbic System:  Sensory Cognitive & Memory          |

&

COPING                          |

    (Focus & Method)

|                                  |

|

= ANS ——–SNS                      |

Increase heart rate |                                  |

|

|                                  |

= CORTISOL ___   ___   ___   ___

                    |

                    |

                    |

_____________________________________________________________________________________________

                                                                                                                       |

                                                                                                                       |

                                                                                  _____________________________________________

 

                                                                                        EMOTIONAL OUTCOMES

• Emotional Intelligence
• Emotional Health

 

Figure 10.  Adaptation from Dorine Felder with permission, (2003), “Schematic overview of the hypothalamic-pituitary-adrenal (HPA) axis. Stress activates the HPA-axis and thereby enhances the secretion of glucocorticoids from the adrenals”. CC BY-SA 4.0 (http://creativecommons.org/licenses/by-sa/4.0)], via Wikimedia Commons. Retrieved December 15, 2016.

Salivary cortisol levels as a major endocrine marker of the HPA activation is in wide use to measure stress responses, allostatic load, allostasis, health, disease and psychiatric pathology. (Turner-Cobb, 2005).   Salivary cortisol sampling is non-invasive and easily, as well as, reliability measured.

The TSST, is another well established standardized protocol for provoking the physiological and emotional stress responses in a laboratory setting for acute stress reactions.   Measurements of TSST cortisol stress responses, in early adulthood (age 23) allows us to examine differences in HPA Axis functioning between former premature and full term infants and prior to later adulthood and later aging process interactions.

Emerging adults, at Age 23, have multiple intrapersonal, interpersonal and stress adaptability reactions to master in order to reach developmental milestones and will require effective emotional intelligence and emotional health.   Societal changes in when emerging adults met the traditional milestone goals and the globally observed increased risk-taking behaviors also warrant examining emerging adults coping styles.

Emerging adults coping styles to deal with stress will not only influence their current emotional health yet will be carried with them into adulthood.   Stress, coping, the cortisol stress response, emotional intelligence, and emotional health are all involved in obtaining developmental milestones at Age 23 for emerging adults.   Results from this study will inform us not solely of current emerging adults’ emotional health status, yet also importantly, if there are any prematurity effects to address that may point to risk for later adult disease.

Secondary Data Analysis This secondary analysis is derived from the dataset of a prospective longitudinal cohort designed study of preterm and full term infants:  Risk & Protection in Trajectories of Preterm Infants:  Birth to Adulthood (Sullivan 2008-2013).   The project received full IRB approval from the University of Rhode Island, Women & Infants Hospital and Rhode Island Hospital.   The study also received a Certificate of Confidentiality from the National Institutes of Health.   The subjects (N = 213) were born at Women & Infant Hospital between 1985-1989 and continually followed at 9 time points in research studies:  1 month, 18 months, 30 months, age 4, age 8, age 12, age 17 and age 23.   The sample retention rate at age 23 was 85% (N = 180).

The original researcher’s theoretical perspective was derived from DOHaD theory and the purpose was to follow preterm (with ranges of gestational age and perinatal morbidity) and full term born infants to age 23 to examine “the impact of prematurity,  medical history, and environments on achievements and deficits during young adulthood (Sullivan, 2008)”.   The original study had 3 specific aims with related hypotheses that focused on determining the effects, relationships, and outcomes of young adults (age 23 years old) in terms of:  health (especially cardiovascular), functional performance, emotional intelligence, executive function, work competence, growth, neurological morbidities and psychological processes.   This secondary analysis adds an in-depth subset analysis of findings related to stress, coping, EI, emotional health and psychiatric disorders of the infant born prematurely at age 23 (Miller & Brewer, 2003).

Compatibility of the primary data with this secondary analysis is not an issue and will still provide information on the differences over time between 23 year olds born pre or full term (Miller & Brewer, 2003; Yea & Niemeier, 1996).   No data collection expenses with be incurred in this secondary analysis study as did in the original longitudinal study.   The first wave of the recruited participants did not change over time and the retention rate (85%) and those lost to attrition at age 23 (15%) remained less than 20% allowing confidence in interpreting the differences between birth groups (Gordis, 2009; Miller & Brewer, 2003; Yea & Niemeier, 1996).

The initial cohort was well characterized and the study procedures and measurements were standardized and well defined.   An “exposed” group of prematurely born and a “non-exposed “ group of full term born infants comprised the two comparison groups and were selected for birth status, neonatal illnesses, gender and wide representation of socioeconomic status.   The original researchers followed protocols without changes in research staff resulting in minimization of personnel effects threatening the internal validity (Miller & Brewer, 2003; Menard, 2007).   As with all longitudinal studies, all factors involved over time, are not captured yet the original design accounted for common confounding variables associated with changes in socioeconomic status between birth and age 23 and gender differences (Menard, 2007).

The two time points used in this secondary analysis add a distinctive period over time and allow for identifying levels of stress adaptability, coping styles, emotional intelligence levels and emotional health status (Singer & Willet, 2003).

Measurements included self-report, interview, observational coding, and standardized protocols used over the course of the longitudinal study were age appropriate, and minimized the effects of the setting and participants’ acclimation (Menard, 2007).

Furthermore, the current researcher had limited involved in the original research nonetheless had access to the principal researchers for clarification, understanding and information related to the study design, data, meaning and decision processes (Miller & Brewer, 2003)

Ethical, confidentiality issues and use of the participants’ identity cleansed data was addressed in the original study consent form at all time points and are not accessible to this researcher (Miller & Brewer, 2003).   Notwithstanding the availability of the original dataset, the selection and in-depth understanding of the dependent variables required extensive analysis.   In addition, modeling analysis of change variables is possible with this design (Singer & Willett, 2003).    What is unique about this secondary analysis is it begins with the identified preterm and full term birth groups before any manifestations of emotional issues and psychiatric disorders develop and focus on the time point of the age 23 year old participants.

Summary

This study examined all dependent variables both alone and in combination that may influence the prematurely born emerging adult in developing emotional health issues.   Before the research plan for this study could be fully developed, a theoretical and concept analysis of the Developmental of Health and Disease Theory, HPA Axis functioning and Emotional Intelligence was done to clarify the concepts and constructs and provide direction in measurement.   Issues in conducting a secondary analysis were reviewed.   In Chapter 3 the secondary analysis and original study description, sample characteristics, methodology, and data analysis plan to investigate any differences in emotional health between the preterm and full term born emerging adult is described.

CHAPTER 3

Methodology

This study is a secondary analysis of a larger longitudinal study aimed at determining the effects of prematurity, risks and protective processes on adult trajectories and health outcomes at age 23. The original study’s theoretical foundation is the Developmental Origins of Health and Disease (DOHaD) and the variable measurements used are consistent with the purpose of this secondary analysis.

This secondary study is a descriptive correlational design of the effect of premature birth and its relationship to stress, coping, emotional intelligence, emotional health and the stress response of neuroendocrine functioning at age 23.   In this chapter, sample and neonatal demographics, self-report measures of stress, coping and emotional intelligence, assessment of emotional health, and timed measures of salivary cortisol through a standardized laboratory social stress paradigm, the Trier Social Stress Test (TSST) are defined. The study procedures and data analysis plans are outlined.

Research Design

The University of Rhode Island, Office of Research Integrity, determined this secondary data analysis, did not require Institutional Review Board (IRB) oversight: # 910108-1 (See Appendix L).

Sample

The infant participants were born at Women’s and Infants Hospital in Rhode Island between 1985-1989, continually followed in a series of research studies supported by the National Institutes of Health (Lester, 1985; McGrath, 1989, 1994, 1998, 2003; Sullivan, 2008 – 2013.   The infant sample included 213 infants in a 5-group design of:  (1) full term healthy infants (FT), (2) preterm infants without neonatal illness (HPT), (3) preterm infants with neonatal medical illnesses (MPT), (4) preterm infants with neonatal neurological illness (NTG) and (5) preterm infants with small for gestational age (SGA) with and without neonatal illness, (Sullivan, 2008-2013). The original racial and ethnic composition of the 213 sampling, consisted of 105 females and 108 males, and was reflective of the population and geographical location in Southeastern New England from 1985 to 1989 (Sullivan, 2008-2013). This original sampling included varied degrees of neonatal morbidity in preterm birth. The study maintained a 97% retention rate from age 4 to 23 years (Sullivan, 2008-2013). At age 23 years, 180 subjects were recruited for participation in the 10^th longitudinal follow-up study with a retention rate from birth of 85%, and 96% from the prior age point of 17 years. Two time points, birth and age 23 years, the 10^th wave of this longitudinal study comprised the sample for this study.

The data (Appendix Measurements) utilized for this study will consist of neonatal data from the original five groups and demographics.   Age 23 years data are, the self-report measures of the Daily Hassles (DHS-R) for stress, Coping Response Inventory (CRI-A) for coping styles, the Bar-On Emotional Quotient Inventory (Bar-On EQi) for emotional intelligence, emotional health status from the Adult Self Report (ASR) and salivary cortisol levels collected according to the established protocol from the Trier Social Stress Test (TSST).

The neonatal sample inclusion and exclusion criteria as well as the characteristics of 5-group design are well described (Sullivan, 2008-2013; Winchester, Sullivan, Roberts & Granger, 2016).  For this study, inclusion criteria included the ability to participate and complete the study protocol and measures.  Any 23-year old participant with severe medical problems, such as cerebral palsy with wheelchair assist, or low intelligence with or without physical limitations was excluded. Females who were pregnant, tested positive using urine pregnancy screening, or <8 weeks post – partum were excluded from the original data collection.  Additionally any 23-year old young adult with positive urine testing for substance use was excluded.

Measurements

The following measurements are proposed for the current study.

Demographic Data

Demographic data included neonatal birth data at age 23 including the Hollingshead Four Factor Index of Social Status (HH), (Holmes & Rahe, 1967).   HH was used to estimate the socioeconomic status of unmarried individuals, female and male heads of households and families (Hollingshead, 1977; Gottfried, 1985).   It is widely used in medicine and public health to differentiate socioeconomic status (Adams & Weakliem, 2011).   The four factors examined consist of:  education, occupation, sex/gender and marital status that are used to derive a mathematical estimate of social status (Hollingshead, 2011).

The occupation scale is scored between 1 and 9, nine being the most highly regarded occupations and one being less highly regarded careers.   Examples of occupations which would receive a score of nine are architects, lawyers and physicians, while occupations that would earn a score of one are janitors, dishwashers and personal attendants.   A seven-point scale is employed to determine educational status.   A seven indicates the completion of a graduate degree or higher while a one represents completion of less than a seventh grade education.   To acquire a social status score the occupation score is multiplied by 5 and the education score is multiplied by 3 (Hollingshead, 1977).   In families with two gainfully employed persons the total score is divided by 2 to obtain a social status score.   The range of computed social status scores (8-66) are then divided into five categories of:  lower (8-19), lower-middle (20-29), middle (30-39), upper-middle (40-54) and upper (55-66), (Hollingshead, 1975).

The Hollingshead Four Factor Index of Social Status is used as a benchmark to compare other measures of socio-economic status (Adams, J., & Weakliem, D., 2011).

The Hollingshead reliability correlates with other indices of SES with ranges of .73-.89 (Cirino, Chin, Sevcik, Wolf, Lovett & Morris (2002).   The social scores achieved are considered the best valid measure of socioeconomic differentiation available (Cassidy, Drotar, Ittenbach, Hottinger, Wray, Wernovsky, Newburger, Mahoney, Mussatto, Cohen, & Marino, 2013; Adams, J., & Weakliem, D., 2011).

Stress

Stress (psychosocial) measurements include everyday and major life experiences as captured by the Daily Hassles Scale-Revised (DHS–R).   The DHS–R, developed by Kanner, Coyne, Schaefer & Lazarus, (1981), focuses on everyday stressors instead of major stress life events and psychological and somatic symptoms (Holm & Holroyd, 1992).  Measures of daily life hassles in 53 items (e.g., time alone, your spouse, your health) on a 4-point scale, ranging from “none or not applicable” to a “great deal.”  The item scores are summed to give a score for the overall severity of hassles, ranging from 0-159.   Internal consistency was .93-.97 over 2 years. Mean day-to-day correlation was .77 and mean monthly correlation was .82.   Daily hassles severity is related to depression, anxiety, PTSD and role stress.   The DHS–R has been previously used in studies with preterm born children (Msall & Park, 2008).

Coping

The Coping Response Inventory-Adult Form (CRI–A) measured coping responses, at Age 23 (Billing & Moss, 1981; Moos, 1995; Moss & Holahan, 2003).

The CRI-A form, developed by Rudolf Moos, is a 48-item self-report inventory measuring cognitive and behavioral responses to stress.   It measures coping orientation focus (approach or avoidance) and methods of coping used (cognitive or behavioral)

The CRI–A is appropriate for use in both healthy adults 18 years of age and older, and those with psychiatric, substance abuse and medical diagnoses.   The instrument may be administered in structured interview or self- report format.   The participants select and describe a recent stressor (within the last year) and use a four-point scale varying from “not at all” to “fairly often” to rate their reliance on each of the 48 coping items.   Eight types of approach and avoidance coping responses are derived from the items.   Approach responses include logical analysis, positive appraisal, seeking guidance and support, and problem solving.  Avoidance responses include cognitive avoidance, acceptance or resignation, seeking alternative awards, and emotional discharge.  Within each of the approach and avoidance responses, the first two scales are representative of cognitive coping strategies and the third and fourth scales of behavioral coping strategies (Moos, 1993).

Additionally, ten appraisal items assess if the stressor was expected, viewed as a threat or challenge, the perceived cause and if resolution has occurred.    Participants respond to questions aimed at examining the context of the stressor using a four-point scale varying from “definitely no” to “definitely yes” (Moos, 1993).

The CRI-A is interpreted using T-scores, with possible T-scores ranging from 20 to 80 + with a mean of 50 and a standard deviation of 10.   A T-score of  < 34 is considerably below average and a score of  > 66 is considerably above average.   Internal consistencies range from 0.61 to 0.74 for males and 0.58 to 0.71 for females.    Cognitive avoidance internal consistency as measured by Cronbach’s alpha is .71 and emotional discharge is .60.   Avoidance coping “across time (1-year stability = .56; 6-year stability = .51) and across life domains (ie., interpersonal, health, and financial stressors), has been found to be stable (Holahan, Moos, Holahan, et al., 2005). “  Test-retests alphas showed moderate stability over 12 months (average  r = 0.45 for males and 0.43 for females).  Content and face validity was built into the CRI-A by formulating definitions of specific domains, preparing items to fit the construct definitions, and selecting items that were conceptually and empirically related to a dimension.   Reliability and validity have been extensively researched and found to be adequate (Moos, 1993 & 2004).

Emotional Intelligence  

The Bar-On Emotional Quotient Inventory (Bar-On EQi) was used to measure the emotional intelligence of all participants. The Bar-On EQi is a self-report measure of emotional and social behaviors and the participants potential to deal with daily environmental demands and pressures (Bar-On, 2002 & 2006;  Sullivan, 2008). It has been used world-wide in a variety of settings with diverse populations and translated into 30 languages since being developed in the 1980’s (Bar-On, 2006). The Bar-On EQi contains 133 items, on a 4-point Likert-like scale (ranging from very seldom to true), resulting in 5 composite scales and 15 subscales (Sullivan, 2008; Bar-On, 2006). The 5 composite scales are: intrapersonal, interpersonal, adaptability, stress management and general mood (Bar-On, 2006)”.  The intrapersonal scale measures self-awareness and self-expression as a composite assertiveness, independence and self-actualization subscales.  The interpersonal scale measures social awareness and interpersonal relationships as a composite score of the social responsibility and interpersonal relationships (“the establishment of mutually satisfying relationships and relating well with others”) scales (Bar-On, 2006).   Stress management measures emotional management and regulation as a composite score of the stress tolerance and impulse control subscales.  The adaptability score measures change management as a composite score of the reality testing, flexibility and problem solving. The general mood composite score measuring self-motivation is derived from the subscales of optimism and happiness. Finally an overall total emotional quotient score is generated.

The 133 items are rated on a five-point Likert-type scale with choices ranging from “very seldom” or “not true” to “very often true” or “true”.   The raw scores are converted into standard scores with a mean of 100 and standard deviation (SD) of 15.

The higher the score the more potential for effective emotional and social functioning and the prediction for meeting daily demands and challenges (Bar-On, 2006).

The Bar-On EQi questionnaire has a built in validity correction factor that adjusts for response bias.   Internal consistencies range from  .69 – .96 with all composite scales correlate highly with the total EQ-I scores (.67 to .93), (Dawda & Hart, 2000).  Test-retest reliability at six months ranges from .72 to .80 (Bar-On, 2006). Ten subscales from a confirmatory factor analysis are the strongest measures of the construct:  self-regard, interpersonal, impulse control, problem solving, emotional self-control, “emotional self-awareness, flexibility, reality testing, stress tolerance, assertiveness and empathy (Bar-On, 2006).”   The remaining 5 subscales are facilitators of emotional and social intelligent behaviors:  independence, self-actualization, social responsibility, optimism and happiness.  The Bar-On EQi has been extensively researched and validated with evidence strongly supporting construct, divergent, discriminant, criterion, and convergent and predictive validity.

Emotional Health

The Adult Self-Report (ASR) was used to determine emotional health. The ASR for ages 18-59 is extensively researched and used widely in research studies for well over twenty years (Achenbach, 2003).   The ASR assesses social competence and behavior problems in adults ages 18-59.   The participant completes the ASR and report their own functioning, problems and substance use.   It provides normative scales for functioning, syndromes, substance use, internalizing, externalizing and total problems (ASEBA, 2011).   The syndrome scales are then profiled to international expertly derived Diagnostic and Statistical Manual of Mental Disorders Fourth Edition Text Revision (DSM IV-TR) oriented scales (Archenbach, 2003).

The ASR (2003) is composed of data relating to friends, spouse or partner, family, job, education and a list of self-descriptors. The descriptor list contains 126 items rated on a 3-point scale consisting of “ not true, somewhat, or sometimes true and very true or often true”. Raw scores are converted in t – scores are categorized into clinical risk screening ratings of normal (t = <59), borderline (t = 60-63), and clinical (t = 64 and above). Classification parameters resulted in true positives (sensitivity) = 80%;  true negative (specificity) = 95%;  false positives = 20%;  false negatives = 5%.

Category scores are related to DSM IV-TR disorder criteria. Internalizing behaviors include anxiety, depression and withdrawn behaviors while externalizing behaviors include aggression, rule breaking and intrusive behaviors. The DSM-oriented problem scales are depressive;  anxiety problems;  somatic;  avoidant personality;  AD/H and antisocial personality.

Reliability is excellent:  1-week test-retest correlations were .80 – .90, 2 years was .69 with none below .71. Internal consistency range was  .78 – .85.  Discriminant validity was demonstrated in the referred and non-referred samples and with DSM IV-TR categories.

Stress Reactivity

The stress reactivity responses were measured by the Trier Social Stress Test (TSST). The TSST, developed by Kirschbaum, is a standardized protocol for inducing stress to subjects in studies and deliberately activates the HPA axis response (Kirschbaum, 2010; Kirschbaum, Pirke & Hellhammer, 1993).   It was used in this study to assess physiological reactivity to what the participant perceives as a challenging situation which has the potential for being negatively judged by others (Kudielka, 2008; Sullivan, 2008 – 2015). In addition to invoking reports of negative mood changes and increased anxiety, the HPA Axis response of endocrine, immunological and other biological markers have been well documented (Stolerman, 2009). It is widely used and involves a well-scripted task the participant is asked to engage in hypothetically while others are evaluating their verbal response. The task consists of a brief preparation period (10 minutes), a test period where the participant delivers a speech for a mock interview (5 minutes) and finally a period of calculating mental arithmetic (5 minutes) in front of an audience.   While completing the tasks the participants are also being video recorded to later assess their behaviors while under stress.   In the original study their behavioral presentation will be evaluated according to pre-established definition levels of:  humor, anxiety, non-verbal but otherwise engaged, defensive, aggressive, scorn, frustration and distraught.

Salivary cortisol samples are collected at the following intervals:  prior to the task (baseline), 15, 30, 45, 60, 75 and 90 minutes after the task has started (Sullivan, 2008-2013).   A mandatory part of the protocol is participants are debriefed after completion of the test (Kirschbaum, 2010).   The TSST has been used with children, adults, healthy subjects and clinical populations.   It is among one of the few tests that assess both behavioral and psychobiological responses to situations that are perceived as uncontrollable and socially threatening (Kudielka, 20008; Dickerson & Kemey, 2004).

HPA Axis Biomarker: Salivary Cortisol Salimetrics, Inc., Expanded Range High Sensitivity (ERHS) Salivary Cortisol Enzymes Immunoassay Kit for research was used to collect samplings during the TSST (Salimetrics, Inc., 2008).   Sensitivity and specificity of the Salivary Cortisol Enzyme Immunoassay is excellent, as well as, a widely used solid technology, and the “gold standard among researchers.   The laboratory procedures include duplicate assays for each cortisol sample so that each saliva sample is tested twice on the assay plate.   This renders two data points per sample and a mean for each sample tested in duplicate.  Duplicate assays minimize the potential of with-in subject variability.

Each assay captures the full range of salivary cortisol levels (0.003 to 3.0 ug/dL) while using only 25 uL of saliva per test (Salimetrics, Inc., 2008).   The lower limit of sensitivity was determined by interpolating the mean minus 2 standard deviations (SDs) for 10 sets of duplicates at 0 ug/dl standard.   For sensitivity, the minimal concentration of cortisol that can be distinguished from 0 is < .003 ug/dL.   The correlation between saliva (Salimetrics ERHS Salivary Cortisol EIA) and serum (Diagnostic Systems Laboratory) was highly significant (r = .91, p = .0001).   During testing, a potential hydrogen  (ph) indicator alerts if cortisol levels are artificially inflated or lowered (< 3.5 or  > 9.0);  possible blood contamination is screened;  and temperature is controlled.   Specificity of antiserum varies but is excellent.   The EIA is accurate in diagnosis of Cushing Syndrome (Raff, Homan, & Skoner, 2003; Lucke, Loucks & Berga, 2007).   Sensitivity/Specificity of 100 %/95 % was reported for salivary cortisol to distinguish diagnosis of Cushing’s syndrome and obesity in an age-matched sample of children (Castro, Elias, Martinelli, Antonnini, Santiago & Moreira, 2000).   Various adult studies exist that measure cortisol levels in a variety of psychological pathologies, emergency situations, and anxiety reduction interventions (Harvey, Nathans, Bandiera & LeBlance, 2010; Rapaport, Schettler & Bresse, 2010).

In the primary study, a careful research protocol was followed.   A foam oral swab was used to collect saliva and placed under the participant’s tongue for 2 minutes. Approximately 1 mL. of saliva saturated the foam swab and then it was inserted into a labeled collection tube and stored in a freezer.   The saliva samples were placed into a cryostorage box with a minimum of 5 lbs. of dry ice and shipped via Federal Express Priority for analysis by Salimetrics, Inc.   Samples were typically submitted to Salimetrics quarterly (every 3 months) but more frequently when necessary.   Shipping instructions according to www.salimetrics.com were followed.

Once analyzed, specimen data were entered into a spreadsheet by Salimetrics, Inc. and reported to the Principal Investigator via e-mail in the form of an Excel file with two values for each saliva sample and the mean in the form of a concentration (e.g., cortisol = 0.652 ug/dL).   Saliva samples were discarded according to Pennsylvania Department of Environment Protection regulations 30 days after testing was completed.   Data from the log were entered into MS Excel by study staff.   The only identifier between the participant and the saliva sample was a predetermined label provided by Salimetrics, Inc., who had no access to data to link the sample to the participant. It was the responsibility of the research staff to link the participant’s unique identification number with the Salimetrics predetermined label.

Summary Data for the proposed study include neonatal and demographic data from infancy and self – report questionnaires, 8 sequential salivary cortisol samples from the TSST standard laboratory protocol from the same participants at age 23 years.   The measures are well developed with good psychometric values.   Questionnaire scoring and subscale information are provided in the Appendix Measurements Procedure Researcher In the study, the researchers were kept blinded to all identifying information which was coded by the primary study staff.   The coded database was not associated with identifying information.   The anonymous complete dataset for the study was provided by the project director.

Recruitment

The infancy sample was recruited during the mother’s post-partum stay or during the infant NICU stay.  The full term infants were identified during the post-partum and in the same time frame as the preterm infants. The criteria for recruitment were neonatal diagnoses, birth weight < 1850 grams (4 pounds), maternal history (no history of mental illness), and English as a primary language.   Parent(s) were then invited to participate by research nurses.   Fewer than 10% of the parent(s) declined participation.

Neonatal Group Criteria was:  (1) full term healthy infants (FT);  (2) preterm infants without illness (healthy preterm group: HPT);  (3) preterm infants with medical illness (bronchopulmonary dysplasia, respiratory distress syndrome, necrotizing enterocolitis and sepsis: MPT);  (4) preterm infants with neurological illness (meningitis, hydrocephalus, and grade 3 or 4 intraventricular hemorrhage:  NPT);  and preterm infants who were small for gestational age (SGA).   Additionally specific protocols for neonatal diagnostic labeling were used.   The range of premature infants in weeks and grams were from 28.3 to 40 weeks gestational age and 1149 (2.5 lbs.) – 3420 (7.5 lbs.) grams in weight.   Socioeconomic status, incorporating both maternal and paternal education and occupation, was measured with the Hollingshead Four Factor Index.

Infants were followed in a series of research studies at ages 18 months, 30 months, 4 years, 8 years, 12 years, and 17 years.   At age 23, the original researchers who had worked with the participants for more than 10 years, made recruitment contact again.   Sample retention was high with 96% of the sample from age 17 years and 85% of the birth sample participating at age 23 years.   The research protocol included a home visit and hospital research laboratory visit.

Demographics

At age 23, demographic information was updated to include name, address, home and cell telephone numbers, e-mail address, social security number, marital status, education level/history, occupation/history, source(s) of financial support, lists of people living in home, pregnancies (for female subjects) and number of children, closet relative information, primary care provider information, and use of medical and community support services.   The participant also filled out his or her own Hollingshead Four Factor Index of Social Status.

The Daily Hassles Scale-Revised (DHS-R), the Coping Response Inventory (CRI-A), the Bar-On EQi and the Adult Self-Report (ASR) were administered to each participant, as part of the larger study protocol of tests and questionnaires, during their first laboratory visit.   Laboratory visits were scheduled according to standardized time periods.   Each participant followed the same schedule or sequencing of questionnaires, food breaks and examinations throughout the specified time period allowing for consistency in protocol and comparison between participants.   This battery of tests was administered after completion of the history and physical in a small private room free from distractions.   The research assistants monitored, timed and administrated the sequencing of tests according to study protocols.

TSST Protocol The Trier Social Stress Test (TSST) followed standardized and extensive protocols relating to timing during the day, timing with other testing and testing protocols.

Levels of cortisol are heightened in the morning and decline throughout the day. To reduce the potential of varied cortisol levels among subjects, administration of the (TSST) was standardized to 1 pm for all participants.   Upon arrival the research nurse (RN) greeted the participant, explained the laboratory protocol, confirmed consent and addressed questions.   To avoid confounders in possible stimulation of the HPA axis, the RN inquired about health-related limitations (colds, flu), physical activity, smoking, medications, large meals, and hormonal treatments within the last 2 hours.

The RN had informed participants of these restrictions when the testing was scheduled. Participants were not informed in great detail about the exact procedure of the TSST.   The RN inquired about the participant’s ideal occupation as this was necessary information for the speech segment of the TSST protocol.   A baseline salivary cortisol sample was collected 20 minutes after arrived (and the time noted) followed by the TSST.

The participants were escorted to an assessment room and told there would be two parts to this segment:  (1) mock job interview and (2) arithmetic problem.  All scripts were followed verbatim.   For the mock job interview, the participant was instructed by the researcher with a scripted standardized comment. The participant was given 10 minutes to prepare for the interview and informed that no notes could be used and they needed to make an impression on the committee (composed of 4 research assistants) who would decide on the acceptance of the application.

The mock job interview was allowed 5 minutes, only one RA would talk and direct the participant and the rest would be silent, with focused eye contact and a serious demeanor that excluded any non-verbal encouragement.    A dummy video camera was set up for the mock interview and the participant was asked to “please step behind this line, and begin your speech.”   All research assistants followed protocols for their roles, behavior, eye contact, facial expressions and speech.   If there was a pause in the speech, then at 20-seconds the participant was told:  “You still have time, please continue.”    Should there be another pause and the participant has nothing to add then the RA will ask a standardized set of questions until the 5-minute time period ended.   At the end of the job interview time the RA announces that the second part of the test will begin and states it had nothing to do with the job interview.

The verbal instructions to the participant for the arithmetic test is also scripted and in part includes:

“We ask you to count backwards to 0 in 13 number steps, starting at 1687. It is important to do it fast and correctly.   Should you miscalculate, you will be told to start again at 1687. Any questions? Please begin.”  

During this time the RA made notes on paper for errors in calculations. If the participant miscalculates, the RA will respond with “Error. 1687.”  After 5-minutes of mental arithmetic, the RA stopped the TSST.

The second cortisol sample was collected 20 minutes after the TSST.   In the interim, the RN conducted the health history interview.   After the collection of the cortisol sample there was a break and then additional questionnaires were administered and collected.   The third cortisol sample was collected after the questionnaires were completed, approximately 30 minutes after the collection of the second sample (50 minutes after the TSST).   The RN then begin the debriefing of the TSST with:

“The first thing I’d like to tell you is that the job interview speech and the arithmetic part were supposed to be difficult.  We did this to elicit a stressful situation for you. When people encounter stressful events, their bodies respond with an accelerated heart rate and the body produces cortisol.  We will be measuring  your cortisol through your saliva.   We were not really recording audio or video for preverbal stress or behavioral analysis.   Also, the judge is actually one of our research assistants.   Let me introduce you to him/her.”

After, the debriefing, a healthy snack was given and further assessment inventories were administered.   The physical examination followed and further salivary cortisol samples were collected at 60, 75 and 90 minutes.

Summary

All measurements representative of the dependent and independent variables to be utilized in this secondary analysis have been reviewed along with the original study procedural protocols.   Use of this data, from the original measurements with good psychometric prosperities, remains a strong fit to the aims of this study despite the longevity of the original study design.

Analysis

All data was analyzed using the IBM SPSS Statistics Standard Grad Pack (SPSS), Version 22 descriptive statistics. The SPSS is used widely in social science research to obtain descriptive statistics (mean, standard deviation, range), graphic depiction, and hypothesis testing (Chi Square, ANOVA, ANCOVA, MANOVA and MANCOVA.

First frequency statistics were performed for each variable. The frequency and pattern of missing data were reviewed.   Missing data imputation was considered when > 10 % data were missing.   Distribution skewness, kurtosis, box-plots were examined and transformations considered when needed.   Correlations were examined between all dependent variables.   Correlations > .80 were considered to behighly correlated and decisions were made to use one variable (FGSE, 2016).   The significance level wasp < = .05.   Analysis of Variance (ANOVA) was used to determine any variations within and between the groups (LoBiondo-Wood & Haber, 2010) in all self-report variables.    AUCg was used with the cortisol endocrine data as a measurement of hormonal output (Fekedulegn, et. al., 2007).   The AUCg is the total area under the plotted curve of all the multivariate measurements and allows for identification of the differences between the single measurements and the distance of these measurements from ground (Spence, Cotton, Underwood & Duncan, 1976).

Measurements from ground (see Figure 11) take into consideration the distance from zero instead of baseline data points and allows for total salivary cortisol response comparisons between participants. (Fekedulegn, et. al., 2007).   Finally, the level of effect power range recommended to reduce any Type I (is effect when there is none) or II (no effect when there is) error at the 0.05 significance criterion is 0.80 – 0.95 (Lakens, 2013).   The study’s’ power was calculated at the 0.80 significance criteria.

Figure 11 Area Under the Curve with Respect to Ground (AUCg)

Figure 11. From:  Catherine Raymond, Marie-France Marin, Anne Hand, Shireen Sindi, Robert-Paul Juster, and Sonia J. Lupien, “Salivary Cortisol Levels and Depressive Symptomatology in Consumers and Nonconsumers of Self-Help Books: A Pilot Study,” Neural Plasticity, vol. 2016, Article ID 3136743, 12 pages, 2016. Figure 4,p 9.  doi: 10.1155/2016/3136743 http://doi.org/10.1155/2016/3136743  Retrieved December 15, 2016.   This is an “open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.”

Aims and Hypotheses

Aim 1: Compare effect of prematurity on stress, coping, emotional intelligence and emotional health.  Compare independent categorical variable of birth groups with the dependent continuous variables:  DHS summed scores ranging from 0-159 for severity of stress, the CRI-A mean scores ranging from 20-80 + with M 50 (SD 10) for avoidance coping, the Bar-On EQI total mean score of 100, (SD 15) for emotional intelligence effectiveness and the ASR mean score of  > 60 for risk to emotional health.

Hypothesis 1. Higher self -reported stress scores, higher use of avoidance coping styles, lower emotional intelligence scores and more emotional health disorders will be found for the adults at age 23 years born prematurely compared to the term-born adults.  Hypothesis 1 will be significant if DHS-R total scores are higher, CRI-A avoidance t-scores are higher, EQi total scores are lower, and ASR scores are higher for the premature groups.

Aim 2:  Compare the salivary cortisol response between premature and full-term born infants in the social stress paradigm of the TSST.   Descriptive statistics including line graphs by groups will be examined first.  Analysis of Variance (ANOVA) will be used to determine differences in salivary cortisol responses as measured by AUCg and at 6 sample times from the TSST paradigm at 15, 30, 45, 60, 75 and 90.

Hypothesis 2.  Adults at age 23 who were born prematurely will have a prolonged stress recovery period for the TSST.   Hypothesis 2 will be significant if the AUCg mean is higher during the stress recovery period for TSST in the prematurely born age 23 year old adult.

Aim 3. Examine the relationship between effect of emotional health and on the stress recovery period of the TSST measured in salivary cortisol.   Three levels of emotional health (normal > 59, borderline 60 – 63, and clinical > 64) as measured by the ASR will be examined for the salivary cortisol samples across 6 sample times.

Descriptive statistics including graph lines by groups will be used followed by ANCOVA where the effects of prematurity and gender will be controlled to understand whether cortisol AUCg and slopes differ due to emotional health.

Hypothesis 3. The stress recovery period for adults at age 23-years with emotional health problems will be prolonged compared to adults without emotional health problems (as determined by ASR t-scores) when prematurity is controlled.

Summary

The measurements used for this secondary analysis are consistent with the aims of the primary study. Measurements and procedures for all dependent and independent variables were reviewed.   The measures have demonstrated good psychometric properties. In the original study, the research team developed and trained on the procedure protocols. Reliability between team members in the protocol were consistently assessed and were >/= 90 %.   The design of this study allows for correlational and multivariate modeling.   The use of neonatal data enables predictive analysis.