Citation
The Effects of demographic and socioeconomic factors on lifestyle behaviors in mothers

Material Information

Title:
The Effects of demographic and socioeconomic factors on lifestyle behaviors in mothers
Creator:
Scheyer, Kathryn
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of Arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Psychology, CU Denver
Degree Disciplines:
Clinical Health Psychology
Committee Chair:
Everhart, Kevin
Committee Members:
Kaplan, Peter
Shaffer, Jonathan
Thayer, Zaneta

Notes

Abstract:
Health disparities are health outcomes that vary based on factors such as race, ethnicity or sociodemographic variables. These differences in health are often associated with lower income and lower educational status which can lead to fewer health care resources and poorer health. The current study will examine one mechanism through which risk factors lead to health disparities—namely, through the lack of engagement in lifestyle promoting behaviors. The current study investigated: (1) the effects of specific socioeconomic and sociodemographic risk factors on the health promoting lifestyle behaviors of mothers, and (2) the psychological factors that influence these relationships. There is research to support a Health Self-Empowerment theoretical model that predicts that some mothers who experience socioeconomic and sociodemographic risk factors will engage in health-promoting lifestyle behaviors because of individual psychological variables. The current study found no significant direct mediation effects of psychological factors on the relationship between specific demographic and socioeconomic factors on maternal lifestyle behaviors. However, significant indirect pathways were found between some variables investigated, including the relationships between economic hardship, educational achievement, personal income, and perception of social status and the pathways of internal health locus of control, health motivation, health self-efficacy and task oriented coping, leading to higher maternal engagement in health promoting lifestyle behaviors

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Kathryn Scheyer. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
THE EFFECTS OF DEMOGRAPHIC AND SOCIOECONOMIC FACTORS ON LIFESTYLE
BEHAVIORS IN MOTHERS by
KATHRYN SCHEYER
B.A., California State University, Long Beach, 2015
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts
Clinical Health Psychology Program
2018


This thesis for the Master of Arts degree by Kathryn Scheyer has been approved for the Clinical Health Psychology Program by
Kevin Everhart, Chair, Advisor Peter Kaplan, Co-Advisor Jonathan Shaffer
Zaneta Thayer


Scheyer, Kathryn B.A., (MA, Clinical Health Psychology Program)
The Effects of Demographic and Socioeconomic Factors on Lifestyle Behaviors in Mothers Thesis directed by Associate Professor, Kevin Everhart
ABSTRACT
Health disparities are health outcomes that vary based on factors such as race, ethnicity or sociodemographic variables. These differences in health are often associated with lower income and lower educational status which can lead to fewer health care resources and poorer health.
The current study will examine one mechanism through which risk factors lead to health disparities—namely, through the lack of engagement in lifestyle promoting behaviors. The current study investigated: (1) the effects of specific socioeconomic and sociodemographic risk factors on the health promoting lifestyle behaviors of mothers, and (2) the psychological factors that influence these relationships. There is research to support a Health Self-Empowerment theoretical model that predicts that some mothers who experience socioeconomic and sociodemographic risk factors will engage in health-promoting lifestyle behaviors because of individual psychological variables. The current study found no significant direct mediation effects of psychological factors on the relationship between specific demographic and socioeconomic factors on maternal lifestyle behaviors. However, significant indirect pathways were found between some variables investigated, including the relationships between economic hardship, educational achievement, personal income, and perception of social status and the pathways of internal health locus of control, health motivation, health self-efficacy and task oriented coping, leading to higher maternal engagement in health promoting lifestyle behaviors.
The form and content of this abstract are approved. I recommend its publication.
Approved: Kevin Everhart


IV
TABLE OF CONTENTS
I. INTRODUCTION..................................................................9
Lifestyle Behavior and Health...............................................9
Demographic and Socioeconomic Factors and Maternal Health...................10
Demographic and Socioeconomic Factors and Child Health.....................12
Lifestyle Behaviors and Maternal/Child Health...............................13
Psychological
Mediators...................................................................15
Research
Hypotheses..................................................................19
II. METHOD.......................................................................21
Participants................................................................21
Design......................................................................21
Procedures..................................................................22
Measures....................................................................22
Demographic questionnaire............................................22
Economic hardship....................................................22
Subjective social status.............................................23
Social support.......................................................24
Active coping........................................................24
Health self-efficacy.................................................25
Locus of control.....................................................26
Maternal depression..................................................26


V
Learned helplessness..................................................27
Lifestyle behaviors...................................................28
Data Analysis................................................................28
Primary Analysis......................................................28
Secondary Analysis....................................................30
III. RESULTS.......................................................................31
Sample
Characteristics..............................................................31
Association among Study Variables............................................32
Primary Analysis......................................................32
Secondary Analysis....................................................37
Post Hoc Analysis.....................................................38
IV. DISCUSSION....................................................................39
Implications.................................................................42
Study
Limitations..................................................................44
REFERENCES..........................................................................47


VI
LIST OF TABLES
TABLE
1. Demographics of Study Sample
2. Descriptive statistics for all variables used in mediation analyses
3. Correlations of all variables
4. Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with psychological mediators
5. Regression of educational achievement on maternal lifestyle behaviors with psychological mediators
6. Regression of personal income on maternal lifestyle behaviors with psychological mediators
7. Regression of family income on maternal lifestyle behaviors with psychological mediators
8. Regression of social support on maternal lifestyle behaviors with psychological mediators
9. Regression of Subjective Social Status (SSS1; current) on maternal lifestyle behaviors with psychological mediators
10. Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with psychological mediators
11. Regression of access to health resources on maternal lifestyle behaviors with psychological mediators
12. Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators


vii
13. Regression of educational achievement on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
14. Regression of personal income on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
15. Regression of family income on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
16. Regression of social support on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
17. Regression of Subjective Social Status (SSS1; current) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
18. Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
19. Regression of access to health resources on maternal lifestyle behaviors with maternal depression and learned helplessness mediators
20. Regression of inability to make ends meet subscale of EHS on maternal lifestyle behaviors with psychological mediators
21. Regression of adjustments and cutbacks subset of EHS on maternal lifestyle behaviors with psychological mediators
22. Regression of not enough money for necessities on maternal lifestyle behaviors with psychological mediators
23. Regression of financial strain subset on maternal lifestyle behaviors with psychological
mediators


VIII
LIST OF FIGURES
FIGURE
1. Diagram of Conceptual framework of three models.
2. Diagram of psychological mediators on relationship between economic hardship and maternal lifestyle behaviors.
3. Diagram of psychological mediators on relationship between educational achievement and maternal lifestyle behaviors.
4. Diagram of psychological mediators on relationship between annual personal income and maternal lifestyle behaviors.
5. Diagram of psychological mediators on relationship between annual family income and maternal lifestyle behaviors.
6. Diagram of psychological mediators on relationship between social support and maternal lifestyle behaviors.
7. Diagram of psychological mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors.
8. Diagram of psychological mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors.
9. Diagram of psychological mediators on relationship between access to health resources and maternal lifestyle behaviors.
10. Diagram of maternal depression and learned helplessness mediators on relationship between economic hardship and maternal lifestyle behaviors.
11. Diagram of maternal depression and learned helplessness mediators on relationship between educational achievement and maternal lifestyle behaviors.


IX
12. Diagram of maternal depression and learned helplessness mediators on relationship between annual personal income and maternal lifestyle behaviors.
13. Diagram of maternal depression and learned helplessness mediators on relationship between annual family income and maternal lifestyle behaviors.
14. Diagram of maternal depression and learned helplessness mediators on relationship between social support and maternal lifestyle behaviors.
15. Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors.
16. Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors.
17. Diagram of maternal depression and learned helplessness mediators on relationship between access to health resources and maternal lifestyle behaviors.
18. Diagram of psychological mediators on relationship between inability to make ends meet subset of EHS and maternal lifestyle behaviors.
19. Diagram of psychological mediators on relationship between adjustments and cutbacks subset of EHS and maternal lifestyle behaviors.
20. Diagram of psychological mediators on relationship between not enough money for necessities of EHS and maternal lifestyle behaviors.
21. Diagram of psychological mediators on relationship between financial strain subset of EHS and maternal lifestyle behaviors.
22. Diagram of SEM model makeup of latent construct “socioeconomic resources” and prediction of maternal lifestyle behaviors.
23. Diagram of SEM of latent construct “socioeconomic resources” prediction of maternal lifestyle behaviors through mediator maternal depression.


X
24. Diagram of SEM of latent construct “socioeconomic resources” prediction of maternal lifestyle behaviors through mediators of binary variables for years of education and race/ethnicity.


1
CHAPTER 1 INTRODUCTION
The Effect of Demographic and Socioeconomic Risk Factors on Life Style Behaviors in Mothers Health disparities are health outcomes that vary based on racial identity, ethnic differences, or sociodemographic variables. Understanding the underlying risk factors for health disparities is a major public health concern. One hypothesis is that health disparities result from differences in stress exposures and environment experiences of disadvantaged populations, as race, socioeconomic status and health have been historically intertwined in the United States (Fiscella & Williams, 2004). In addition, lower income and education often lead to less access to health care resources and the development of poor health. The current study examined an additional mechanism through which health disparities may occur in contemporary society: differential engagement in lifestyle promoting behaviors. I specifically evaluated the relationships among sociodemographic and socioeconomic factors and engagement in health promoting lifestyle behaviors by mothers, and (2) the psychological factors that influence these relationships.
Lifestyle Behavior and Health
Health-promoting behaviors, such as exercise, practicing good nutrition and utilizing stress management techniques, have been linked to better mental and physical health outcomes across populations. Physical activity has been found to be preventative for disability in aging populations (Buford, Anton, Clark, Higgins, & Cooke 2014), beneficial for improving overall health (Mitchell & Barlow, 2011) and preventing and/or treating numerous medical conditions (Walsh, 2011). A five-year review on exercise and quality of life (QOL) found a general consensus in the literature that higher levels of exercise result in higher QOL in both healthy


2
individuals and those with specific ailments, such as those found in aging populations (Mitchell & Barlow, 2011). Nutrition may also be considered a modifiable risk factor for some psychopathology (Low Dog, 2010), with implications for both physical and mental health. Overall, improvements in eating habits and lifestyle, including dieting and physical activity, are significantly positively associated with long-term improvements in mental health, such as depression, anxiety, self-esteem and negative affect. Higher levels of physical activity and healthy eating habits have resulted in improved QOL and improved psychological health (including a reduction in depression and anxiety and improvements in self-esteem) for adult populations, children and adolescents (Biddle & Asare, 2011; Schaefer & Magnuson, 2014). Healthy eating and engaging in physical activity enhance cognitive functioning, academic achievement among adults, children and adolescents and reduce age-related memory loss and the risk of dementia in elderly populations (Biddle & Asare, 2011; Walsh, 2011).
Additionally, stress management techniques, including mindfulness-based stress reduction decreases stress among populations with multiple physical ailments and psychological difficulties, including medical symptoms, pain, physical impairment, depression, and anxiety (Grossman, Niemann, Schmidt & Walach 2004). A review article on lifestyle and mental health describes therapeutic lifestyle changes (TLC), including exercise, diet and nutrition, recreation, relaxation and stress management and religious or spiritual involvement, as a central focus in overall mental and physical health (Walsh, 2011).
Demographic and Socioeconomic Factors and Maternal Health
Given that previous research provides a link between engaging in health promoting lifestyle behaviors and better mental and physical health (Prather, Spitznagle, & Hunt, 2012), it is important to understand the relationships among risk factors and the health of mothers and


3
infants. Previous research on the health of mother and infant populations has shown that early life exposure to social disadvantages or unfavorable health conditions have potential long-lasting effects. Maternal chronic stress, which can be shaped by socioeconomic and sociodemographic factors such as low socioeconomic status (SES), low social and partner support, low educational achievement, and low access to resources, significantly jeopardize the health of both mother and baby (Ghosh, Willhelm, Dunkel-Schetter, Lombardi, & Ritz, 2010; Lancaster, et al., 2010; Lueken et al., 2008; Matijasevich et al., 2010; Premji, 2014; Thayer & Kuzawa 2014). Matijasevich et al. (2010) assessed health outcomes of socioeconomic factors that contribute to health disparities through multiple birth cohort studies in Brazil and the United Kingdom. They found that mothers from poorer and less educated backgrounds had more adverse health behaviors and outcomes, such as smoking during pregnancy, preterm birth and shortened duration of breastfeeding practices (breastfed infant for less than 3 months). Sociodemographic factors such as income, educational attainment and partner status have also been found to be predictive factors of depression symptoms during and after pregnancy. Social factors, specifically social support and marital relationships, also influence the perinatal mental health of women (Premji, 2014) and a decrease in social support is negatively associated with maternal health problems. Through the large scale Fragile Families and Child Wellbeing Study (N = 12,140), Harknett & Hartnett (2011) found that less social support and lack of personal safety nets through social networks were predicted by lower income and further, mothers with poor physical and mental health (higher rates of depression) were less likely to have financial, housing or child-care support available to them.
In addition to experiencing prenatal stress and health problems, mothers from relatively low-income backgrounds have historically received poorer maternal care. Abel (1996) studied


4
the relationship of socioeconomic factors such as ethnicity (African American versus White/ Caucasian), education (less than a high school education versus high school education) and marital status (married versus unmarried) with the receipt of adequate prenatal care among 11,936 birth cases and found that these risk factors significantly predicted inadequate prenatal care. Lower income, fewer years of education and residential property status predicted higher scores on the Edinburg Postnatal Depression Scale (EPDS) and that partner status (i.e., not having a partner) and family support were most predictive of depressive symptomology in mothers (Hein et al., 2014; Lancaster et al., 2010).
Demographic and Socioeconomic Factors and Child Health
Maternal and child health are interrelated. Poor maternal health and elevated maternal stress, due to low-income and other sociodemographic risk factors, is predictive of both adverse birth outcomes (Dominguez, 2011; Gennaro & Hennessy, 2003; Hobel, Dunkel-Schetter,
Roesch, Castro & Arora, 1999) and poorer child health outcomes (Minkovitz, O’Campo, Chen,
& Grason, 2002; Hardie & Landale, 2013). Studies have shown a link between prenatal maternal stress, which seems to be pronounced in low SES and ethnic minority families (Luecken et al., 2010) and adverse infant health outcomes, including preterm birth (Dole & Mollborn, 2003), low birth weight (Dennis et al., 2013; Parker, Schoendorf, & Kiely, 1994), and infant mortality (James, 1993). Consequently, inadequate prenatal care predicts low birth weight and infant mortality (Gortmarker, 1979). There is also research that supports the relationship between poverty (income and maternal educational achievement), child stress physiology, and child health and development outcomes after infancy.
Socioeconomic status levels and environmental threat and hostility, which contribute to maternal stress, may have long-lasting effects on the developing hypothalamic-pituitary adrenal


5
(HPA) regulating system. Children who grow up in hostile or threatening environments may develop a highly sensitive HPA system in response to environmental stress and the absence of socioeconomic resources (e.g., social support) (Everhart & Emde, 2006). Physical and psychosocial stressors, such as housing quality and violence were found to be associated with heightened cardiovascular and neuroendocrine levels and cumulative stress in early childhood (Evans, 2003; Everhart & Emde, 2006). Additionally, cultural and environmental context, such as low SES and maternal stress significantly predict higher infant cortisol reactivity after birth (Thayer et al. 2014) and higher salivary cortisol levels in infants and at 4.5 years of age (Clearfield et al., 2014; Essex, Klein, Cho & Kalin, 2002). Furthermore, socioeconomic influence on HPA regulatory systems may be responsible for dysregulation in early life and child psychopathology (Everhart & Emde, 2006). Elevated salivary cortisol and reactivity has potential long-term effects on child biology and health through a variety of psychological and physical health problems such as anxiety disorders, depression, somatic complaints, aggression, attention problems, cardiovascular disease, respiratory disease and some types of cancer (McEwan, 2007; Miller et al., 2009; Santiago, Wadsworth & Stump, 2011).
Lifestyle Behaviors and Maternal/Child Health
Health promoting behaviors are particularly important in the area of maternal and infant health, as previous research has established a relationship between healthy lifestyle behaviors and the mental and physical health of both mother and infant. Prather et al., (2012) for instance, reviewed previous studies showing that exercise and appropriate nutrition are important contributors to maternal and infant physical and psychological health. Their review identifies maternal benefits of exercise, including improved cardiovascular function, lower risk for gestational diabetes, improved strength and lean muscle mass, improved sense of well-being, and


6
enhanced sleep. It further found that exercise has anti-depressant effects, improves self-esteem and reduces symptoms of anxiety and depression during pregnancy. Maternal exercise during pregnancy also affects infant health, resulting in higher birth weights and potentially improved neurodevelopment (improved orientation, ability to self-sooth in neonates and higher general intelligence and oral language scores at age five; Prather et al., 2012). Maternal nutrition, including appropriate diet with sufficient vitamin D and weight management, is also an important factor that plays a role in shaping the health of mothers and infants (Prather et al.,
2012; Uriu-Adams, Obican & Keen, 2013).
Relaxation techniques and other stress management interventions positively affect maternal health. Improvements in health include improved emotional state during pregnancy, improved pregnancy outcomes (e.g., fewer admissions to the hospital, fewer postpartum complications), improved fetal and neonatal outcomes, including a reduction in fetal heart rate and fetal motor activity, a reduction in maternal physiological and endocrine measures (Fink, Urech, Cavelti, & Adler, 2012) and reduced postpartum stress in general (Song, Kim & Ahn, 2015).
Stress researchers have found that individuals from low-income or low-educational backgrounds have the highest rates of morbidity, disability, mortality, psychological distress and mental disorders compared to those from more advantageous socioeconomic backgrounds (Thoits, 2010). Socioeconomic and sociodemographic risk factors affect a variety of lifestyle behaviors that promote physical and mental health and specific resources can help individuals deal with demands and cumulative psychosocial stressors. Mothers from low-income backgrounds may not possess the coping resources, knowledge, money, or access to healthcare that mothers from higher SES backgrounds have through which to secure good health (Thoit et


7
al., 2010; Roux et al., 2013). Indeed, previous research has shown that education and income are positively associated with health promotion practices in general (practicing good nutrition, stress management; Duffy, 1997; Walker, Sechrist, & Pender 1987) and negatively associated with the risk of weight gain (Kahn, Williamson, & Stevens 1991). Additionally, Duffy et al. (1997) found that specific demographic factors, such as educational level, were associated with improved frequency of engaging in health promoting activities including exercise, nutrition, stress management, interpersonal support, health responsibility and self-actualization in mothers.
Other literature also suggests that mothers from low-income and minority backgrounds have less access to economic resources and knowledge that promotes healthy lifestyle behaviors such as exercise and stress management. In underserved populations (such as populations of low SES, racial or ethnic minorities or non-English speaking; Weitz, Freund & Wright, 2001) access to maternal health resources and factors such as time and money serve as barriers for women to engage in these specific health-promoting behaviors (Duffy et al., 1997; Gazmararian, Adams, & Pamuk 1995). Gazmararian et al. (1995) found specific sociodemographic risk factors such as education level, poverty and Medicaid status, predict some maternal lifestyle behaviors such as smoking. Specific health behaviors detrimental to health outcomes, known as health-risk behaviors (e.g., smoking), are also more prevalent in low-income populations, specifically among those with low SES and educational status that are particularly harmful to maternal and infant health outcomes (Gazmararian et al., 1995; Lantz et al., 2001).
Psychological Mediators
Given findings that show a consistent relationship between demographic and socioeconomic risk factors and maternal and infant health, and the positive relationship between engagement in health-promoting lifestyle behaviors and overall health, the current study aimed


8
to understand the specific relationship between these socioeconomic and sociodemographic factors and lifestyle behaviors in mothers and what factors may mediate this relationship. I hypothesize that mothers’ individual attitudes about self and health will mediate the association of socioeconomic and sociodemographic factors with the engagement in health-promoting lifestyle behaviors.
Specific psychological variables may mediate the relationship between risk factors and lifestyle behaviors and explain why some women engage in health-promoting lifestyle behaviors. For instance, research has shown that psychological variables serve as predictive factors of health-promoting lifestyle behaviors. A review by Matthews, Gallo and Taylor in 2010 proposes a framework demonstrating how individuals living in low SES conditions have a “smaller bank” of resources to cope with stressful events. According to this framework, limited coping resources lead to poorer health outcomes. According to these authors, elevated negative emotions and cognitions lead to psychological pathways causing poor long-term health. Furthermore, individual attitudes of mothers, which may include attitudes toward healthful eating (Clarke, Freeland-Graves, Klohe-Leman & Bohman, 2007; Jordan et al., 2008), mood self-efficacy (Chang, Brown & Baumann, 2011), health motivation (Jayanti & Burns, 1998) and motivation to exercise, exercise self-efficacy and overall self-efficacy (Clarke et al., 2007; Mailey & McAuley, 2014), serve as predictive factors of maternal engagement in various health-promoting lifestyle behaviors. Matthews, Gallo and Taylor (2010) also identify various positive psychological resources, such as self-esteem and mastery, in the association between SES and health, but recognize a need to conduct further research on positive emotions that lead to intermediate, mediating psychological pathways and health.


9
According to Bandura’s Social Cognitive Theory (SCT; Bandura, 1986) cognitive/personal (self-efficacy) and social/environmental variables (demographic and socioeconomic variables) are important determinants of health behavior. SCT is often used to understand the underlying causes, or mechanisms, in the occurrence of health-promoting behaviors (Tucker, Butler, Loyuk, Desmond & Surrency, 2009). Health Self-Empowerment (HSE) Theory, a theory that builds upon SCT, may help to further explain differences in health-promoting behaviors among members of populations who are dealing with the same social/environmental barriers that SCT cannot account for. HSE theory acknowledges the influence of environmental stressors, such as poverty, on health behaviors but asserts that key psychological variables and individual attitudes in minority populations and low-income communities can significantly influence health behaviors. The current review of literature finds only one study which has examined the relationship between HSE theory, including multiple and specific psychological variables and health-promoting lifestyle behaviors. Tucker et al. (2009) examined this relationship among low-income African American mothers and white mothers of chronically ill children. Active coping, health self-efficacy, and health motivation were among the psychological variables that were found to significantly affect mothers’ healthy diet, exercising, using stress management practices, and engaging in health responsibility behaviors (Abusabha & Achterberg, 1997; Tucker et al., 2009).
Additionally, because a key facet of engaging in health-promoting lifestyle behaviors is a pattern of self-initiated actions and perceptions about individual health behavior and health wellness (Walker et al., 1987), locus of control may also be a psychological mechanism through which women engage in specific lifestyle behaviors. Rotter (1954) developed the idea of individuals having either internal or external locus of control. Those with higher internal locus of


10
control view events as resulting from their own actions, while those on the other end of the continuum, with higher external locus of control view events as being under the control of external factors. A multidimensional health locus of control construct has been used that assesses the specific role of beliefs in the context of health behavior, health outcomes and health care, and can help to contribute to the understanding of how perceived self-efficacy and locus of control may be involved in the prediction of lifestyle behaviors (Abusabha et al., 1997).
Maternal depression is negatively correlated with health behaviors (Haller, Knisely, Dawson & Schnoll, 1993; Lindgren, 2001; Walker, Cooney, & Riggs, 1999). Thus, the presence of maternal depression is hypothesized to lead to a reduced likelihood of healthy behaviors. In the current study, we measured maternal depression. In addition, we focused on one cognitive aspect of depression: learned helplessness. Learned helplessness is a related concept to locus of control. In Seligman’s theories of learned helplessness (Hiroto & Seligman, 1975) the central idea is that if people feel they have no control over future outcomes, they are less likely to seek solutions to their problems. This chronic discovery of a loss of control leads to the concept of learned helplessness, a cognitive component of depression, and may be another mechanism that affects the relationship between environmental factors and lifestyle behaviors (Seligman, 1974). Furthermore, these concepts are integrated into Bandura’s SCT, from which HSE theory stems, in that self-efficacy influences the expectations people hold about their abilities to control future outcomes and accomplish goals. Active coping and health motivation may positively correlate with self-efficacy and internal locus of control and negatively correlate with learned helplessness. The current study measured task-oriented coping, health self-efficacy, health motivation, and locus of control as psychological variables that may cause women to engage in health promoting behaviors despite environmental, social and economic stress. Using primarily


11
HSE, the current study builds a conceptual framework (see Figure 1) to help show this framework. The current study is the first of my knowledge that investigates multiple mediating variables that may influence the relationship between demographic and socioeconomic factors on lifestyle behaviors in a sample of mothers as a means of assessing greater health disparity implications.
Research Hypotheses
The current study assessed a multidimensional approach of sociodemographic and socioeconomic factors with one approach being through the use of the Economic Hardship Scale (EHS; Barrera et al., 2001). The EHS is a subjective measure of economic hardship that provides information about the psychological sense of disparity between needs and resources. Another subjective indicator of socioeconomic position, the MacArthur Scale of Subjective Social Status, was used to assess each mothers’ perception of social status that reflects the relative perception that individuals have of their place in the social hierarchy. Subjective aspects of social position contribute to our understanding of the relationship between socioeconomic risk factors and health problems (Wilkinson, 1997) as socioeconomic gradient and subjective experience of social status of individuals may be a better predictor of health than objective variables of social status, such as education or material wealth (Everhart & Emde, 2006). Each measure being used helps to understand unique dimensions of subjective life experience, including demographic aspects (access to health resources), perceptions of social status, social support, and psychological sense of hardship.
The current study identified three separate models that represent this multidimensional approach. The current study examined psychological variables that serve as potential mediators of the relationship between demographic and socioeconomic factors, and maternal engagement


12
in health-promoting lifestyle behaviors. Figure 1 depicts a visual representation of the conceptual framework proposed in the study.
Model 1
The first model of the study examined the influence of a subjective perspective of economic hardship on lifestyle behaviors. The current study is the first to my knowledge to examine the influence of a subjective perspective of sociodemographic and socioeconomic factors on maternal engagement in health promoting lifestyle behaviors.
Model 2
Secondly, because there is an abundance of evidence that socioeconomic status, educational achievement and social support significantly affect lifestyle behaviors and overall maternal and child health, the secondary model of the study was to measure these three variables. The second model of the study also measured individual perception of social status to further understand another dimension of perception of hardship.
Model 3
Lastly, access to resources was a third model of the study since research shows that the above independent variables, such as SES and educational achievement, may affect receipt and access to health resources and care.
It was expected that the relationship between socioeconomic risk factors and engagement in health-promoting lifestyle behaviors would be significantly influenced by psychological mediating factors of mothers who come from underserved backgrounds. I predicted that despite coming from low-income, high-stress environments and being exposed to socioeconomic risk factors, some mothers, because of individual psychological variables, such as health-motivation, locus of control and self-efficacy, will engage in health-promoting lifestyle behaviors.


13
CHAPTER 2 Method
Participants
Sample selection and eligibility criteria. Mothers were recruited from an advertisement posted on our laboratory’s Facebook page to do an online survey. The advertisement described the project as a study “about maternal background and environment, lifestyle behaviors and individual attitudes.” Participants were not provided with any monetary compensation for their participation. Informed consent was obtained from a brief initial section through the online survey.
Design
This study featured a cross-sectional design with a mediation analysis. The independent variable was defined as demographic and socioeconomic factors and included: three sociodemographic variables (income, educational achievement level and level of access to resources), a total score for the experience of economic hardship, a total score for perceived social support, and two total scores for subjective social status (current and future perceptions). The dependent variable was defined as maternal engagement in health-promoting lifestyle behaviors and includes a total score of health-promoting lifestyle behaviors. The mediating variables were defined as the underlying psychological variables that are hypothesized to influence mothers from low-income backgrounds to engage in health promoting behaviors and include health self-efficacy, health motivation, task-oriented coping, locus of control and depression. These variables and their measurement are described in greater detail in the
‘Measures’ section.


14
Procedures
Participants included 159 mothers who were asked to fill out a Qualtrics online survey through a link that was posted on the laboratory’s Facebook page. The survey took approximately 30-40 minutes to complete and included measures in a randomized order. Randomization was conducted by presenting question blocks in random order using the Randomizer feature on Qualtrics. The scales listed below were converted to an online format which may affect the reliability of each scale. Reliability testing was done to assess any changes in internal reliabilities of the scales. With the exception of the total score for access to health resources (.59) and the Subjective Social Status Scale (.68), Cronbach’s alpha was at or above .70 for all of the scales converted into online form and used in the current study. Participants had the option to leave their name and email and/or phone number if they were interested in participating in future studies.
Measures
Demographic Questionnaire. In order to assess objective measures of demographic and socioeconomic factors, the current study included a demographic questionnaire. The questionnaire asked about age, ethnicity, race, marital/ relationship status, personal and family income, educational achievement and employment status, and access to medical and health resources.
Economic Hardship. In order to assess a multidimensional perspective of socioeconomic and sociodemographic risk factors, the current study used the Economic Hardship Scale created by Barrera et al. (2001). The Economic Hardship Scale is a 20-item self-report


15
measure and is a rated on a 5-point Likert type scale ranging from “almost never” to “almost always” or “strongly agree” to “strongly disagree”. The construct defines economic hardship as (a) the inability to afford specific necessities for living, (b) a general sense that financial obligations exceed the family’s ability to meet them, (c) behavioral attempts to reduce expenses or generate more income, and (d) hopelessness that the future will bring a brighter financial outlook. The construct shows reliability in assessment of perceived economic hardship among multiple ethnic groups including African American, European American, and Mexican American families who reside in urban areas (Barrera et al., 2001). Results of reliability analysis from the current study are in line with previous research. Internal reliabilities (Cronbach’s alpha) are provided for four subsets includes in the scale: Economic Adjustments and Cutbacks (.79) Not Enough Money for Necessities (.95), Inability to Make Ends Meet (.86) and Financial Strain (.80) Examples of questions asked on this scale include “Tell me how much you agree or disagree with these statements: We had enough money to afford the kind of car we need; In the past three months, we changed food shopping or eating habits a lot to save money.” This measure can objectively be obtained through a demographic questionnaire but also provides a subjective sense of hardship, individual need and personal and financial struggle.
Subjective Social Status. The MacArthur Scale of Subjective Social Status was administered as a measure of perceived social status across different levels of socioeconomic status (Adler & Stewart, 2007). This measurement features a pictorial “social ladder,” and individuals are asked to place an “X” on the rung on which they feel they stand currently. Reliability testing has shown good internal reliability in previous research (Giatti, Valle Camelo, Rodriquez, & Barreta, 2012), as well as in the current study. The current study also asked individuals to place an “X” on the rung on which they feel they will stand 5 years from now.


16
Internal reliability testing shows the “social ladder” has a Cronbach’s alpha of .68, between the two questions, current and future. This scale assesses common sense of social status across different SES and can include perceptions of community, familial and societal social statuses.
Social Support. The Multidimensional Scale of Perceived Social Support is a 12-item self-report measure of subjectively assessed social support, and is rated on a 5-point Likert-type scale ranging from “strongly disagree” to “strongly agree.” The scale contains a total score (Cronbach’s alpha of .84) and three subscales: Family (Cronbach’s alpha of .91), Friends (.95) and Significant Other (.91). Example statements include “There is a special person with whom I can share my joys and sorrows” and “I get the emotional help and support I need from my family”. Previous research has also shown good internal reliability (Zimet, Powell, Farley, Werkman, & Berkoff, 1990). This measurement allowed us to assess individuals’ perception of how much they receive outside social support that can include family, friends and significant others.
Active Coping. The current study used the Coping Inventory for Stressful Situations Short Form (CISS-SF; Cosway, Endler, Sadler, & Deary, 2000) which is a 21-item self-report inventory created to measure task-oriented (e.g. “Focus on the problem and see how I can solve it”), emotion-oriented (e.g. “Blame myself for being too emotional about the situation”) and avoidance-oriented (e.g. “Take time off and get away from the situation”) coping. This scale uses a 5-point Likert scale ranging from “Not at All” to “Very Much”. Internal reliabilities (Cronbach’s alpha) for three subsets of the scale are: task-oriented (.84), emotional (.84) and avoidant (.65), and (.74) for the total score. The CISS-SF scale allowed us to assess coping styles of mothers and day-to-day coping of stressful situations that play a role in physical and psychological well-being.


17
Health Self-Efficacy. The School Health Efficacy Questionnaire (SHEQ) is a 43-item inventory that measures health self-efficacy—that is, the degree of confidence individuals have in their ability to perform specific health behaviors that contribute to their overall physical and mental well-being. Most health behaviors measured by this scale are not specific to school, making it possible to use this scale with adults with minor revisions. The current study used 35 out of 43 questions that applied to the target population, taking out questions regarding school health behaviors. The SHEQ has a seventh-grade reading level, allowing measurement in groups of individuals from low-educational achievement backgrounds. SHEQ responses are reported using a 5-point Likert scale ranging from “very little” to “quite a lot.” Reliability testing show a Cronbach’s alpha of .91 for the total scale and is broken down into two subscales: Physical Health (example item: “Knowing when I am getting sick;” Cronbach’s alpha of .80) and Mental Health (example item: “Avoiding worry about trivial things;” Cronbach’s alpha of .83)
Reliability testing has also shown good internal consistency in previous research (Froman & Owen, 1991). Because of the population being studied, the level of understanding of the SHEQ helped us to understand the self-efficacy perceptions that motivate health-promoting lifestyle behaviors in mothers.
Health-Motivation. The Value on Health Scale (VHS) is a 5-item self-report measure designed to assess value on, preference for, or personal importance of several aspects of health: fitness of being in good physical condition, a sense of energy or vigor, endurance or stamina, maintaining an appropriate weight, and resistance to illness. The measure uses a 4-point Likert scale ranging from “not important” at all” “to very important.” Example questions include: “How important is it to you to be in good shape and to feel physically fit?” and “How important is it to you to feel you have plenty of energy for the way you’d like to live your life?” The VHS has an


18
alpha reliability of .84 and has good internal reliability according to previous studies (Costa, lessor, & Donovan, 1989). This scale was used to better understand how values placed on health are related to greater involvement in health-enhancing behaviors, and how psychosocial and behavioral characteristics may explain differences in health behaviors.
Locus of Control. The current study used the Multidimensional Health Locus of Control Scale (MHLOC) to assess locus of control regarding health behaviors (Wallston, Wallston & DeVellis, 1978). The MHLOC is an 18-item self-report measure that assesses individuals’ belief about what influences health using a 6-point Likert scale ranging from “strongly agree” to “strongly disagree.” The scale assesses three dimensions of locus of control (Cronbach’s alpha for the total scale is .70): (1) internal belief (.75): “My health is influenced by my own choices and behaviors,” (2) chance belief (.74): “My health is influenced by chance or fate and neither me nor my doctor have much influence on it,” and (3) powerful others belief (.73): “My health is dependent on the competence of my doctor.” Reliability analyses in previous research shows adequate internal reliability (Thompson, Butcher & Berensen, 1987). The MHLOC offers an understanding of locus of control in reference to perceived self-efficacy and its involvement in health behaviors.
Maternal Depression. The Edinburgh Postnatal Depression Scale (EPDS; Cox et al., 1987) was used to assess symptoms of maternal depression. The EPDS is a 10-item self-rating scale designed to detect postnatal depression and uses a 4-point Likert scale ranging from “As much as I always could” to “Not at all” and “Yes, most of the time” to “No, never”. The total possible score ranges from 0 to 30 and indicates the possible degree of postnatal depression experienced during the past week. Higher scores indicate more severe maternal depression. Previous research has found the best cutoff scores to be between 9.5 and 12.5 (Guedeney &


19
Fermanian, 1998), and multiple studies have used a cutoff score of 9/10 to indicate the incidence of severe depressive symptoms (Jadresic, Araya & Jara, 1995). The current study used the PEDS as a continuous score from 0-30, with higher values representing higher rates of depressive symptomology. The EPDS has been used with a variety of samples and has shown good psychometric properties (Dayan et al., 2006; Eberhard-Gran, Eskild, Tambs, Opjordsmoen & Samuelsen, 2001). Reliability studies confirm good internal consistency of the EPDS in postnatal depression-symptomology evaluation (Bergink et al., 2011; Pop, Komproe & van Son, 1992; Teissedre & Chabrol, 2004). Cronbach’s alpha for the total scale is .89 in the current sample. Example statements include “I have felt happy” and “I have been so unhappy that I have been crying”.
Learned Helplessness. Five questions from the Learned Helplessness Scale (LHS; Quinless & Nelson, 1988) were used to assess learned helplessness and combined with questions from the EPDS. The original LHS scale is a 20-item, self-rating Likert type scale ranging from “Strongly agree” to “Strongly disagree”. The total possible score of the original scale ranges from 20 to 80, with higher scores suggesting greater helplessness due to the perception that events are beyond the respondent’s control. The five questions that were used in the current study to assess for learned helplessness as a part of the maternal depression subscale were: “In general- No matter how much energy I put into a task, I feel I have no control over the outcome”, “In general- If I complete a task successfully, it is probably because I become lucky”, “In general- If feel that my own inability to solve problems is the cause of my failures”, “In general-When something doesn’t turn out the way I planned, I know it is because I didn’t have the ability to start with” and “In general- Other people have more control over their success and/ or failure than I do”. Reliability coefficient (Cronbach’s alpha) of .87 of the shortened version of the LHS


20
used in the current study. Reliability analyses on the 20-item LHS also have confirmed good internal consistency (Quinless & Nelson, 1988).
Lifestyle Behaviors. The Health-Promoting Lifestyle Behaviors Profile (HPLP-II) was used as a multidimensional assessment of health-promoting behavior patterns that include self-initiated actions and perceptions that influence mothers’ engagements in health-promoting behaviors. The HPLP-II is a 52-item behavior rating scale that assesses the frequency of self-reported health-promoting behaviors in the domains of health responsibility, physical activity, nutrition, spirituality, interpersonal relations and stress management. There are six dimensions to the HPLP-II scale: Self-Actualization, Health Responsibility, Exercise, Nutrition, Internal Support and Stress Management. The reliability coefficient (Cronbach’s alpha) for the total scale is .95. Reliability analyses confirm good internal consistencies in previous research (Walker, Noble, Sechrist, Richert, & Pender, 1987). Example questions include “I choose a diet low in fat, saturated fat and cholesterol,” “I follow a planned exercise program,” “I take some time for relaxation each day,” and “I accept those things in my life which I cannot change”.
Data Analyses
Primary Analysis
I hypothesized that the independent variables of economic hardship, subjective social status, income, educational achievement and access to resources, and social support influence the mediator variables, psychological variables including health self-efficacy, health motivation, task-oriented coping and health internal health locus of control, which in turn influence the dependent variable, health-promoting lifestyle behaviors. In the statistical analysis, we tested this using a multiple mediation analysis considering our hypotheses that multiple mechanisms (mediators) are acting on this phenomena at once. The current study first used a serial multiple


21
mediator model regression analysis, assuming the mediators are linked together in a causal chain (Hayes, 2013). In a serial multiple mediator model, the total effect of X on Y partitions into direct and indirect components. The analysis includes obtaining the direct effect c1, which is interpreted as the estimated difference in Y between two cases that differ by one unit on X but are equal on all mediators. The serial multiple mediator model contains indirect effects estimated as products of regression coefficients linking X to Y. The indirect effects, of which there are many due to the number of mediators in the model, are constructed by multiplying the weights corresponding to each step in an indirect pathway. These indirect effects are found in the PROCESS output along with 90% bias-corrected bootstrap confidence intervals based on 5,000 bootstrap samples. Both a 95% and a 90% confidence interval were used in the analyses and no significant differences were found between the two. The numbers reported are using a 90% confidence interval in order to have a larger margin of error. Indirect effects are all interpreted as the estimated difference in Y between two cases that differ by one unit on X through the causal sequence from X to mediators to Y. In order to express indirect effects in terms of the difference in standard deviations in Y between two cases that differ by one standard deviation in X, a completely standardized effect size was used and denoted as ccs. Before testing the multiple mediators in the model, the first step in the analysis is to confirm that at least two or more of the proposed mediators are correlated with each other, making the model a serial multiple mediator model (Hayes, 2013). Due to the significant correlations between the mediators (shown in Table 3), we used a serial multiple mediator model, or Model 6, in the PROCESS Macro. Since the serial multiple mediator model only allows for four mediators in the model at one time, health-related psychological mediators (internal health locus of control, task-oriented coping, health


22
self-efficacy and health motivation) were run in the first step of analyses, while maternal depression and learned helplessness were run in the second step of analyses.
Secondary Analysis
A secondary analysis was also done using structural equation modeling (SEM) to test the hypothesized relationships between demographic and socioeconomic factors, maternal lifestyle behaviors and maternal depression, one of the proposed mediators. SEM has been described as a combination of confirmatory factory analysis and multiple regression (EUlman, 2001). SEM explores the possibility of relationships among latent variables and encompasses two components: (a) a measurement model and (b) a structural model that allows the ability to test a full theoretical model among variables of interest that can be latent factors or directly measured. The measurement model of SEM is the confirmatory factor analysis and depicts the pattern of observed variables for those latent constructs in the hypothesized model. The structural model comprises the other component of linear structural modeling and displays the interrelations among latent constructs and observable variables in the proposed model as a succession of structural equations.
I performed an SEM analysis based on data from 288 mothers using the statistical modeling program Mplus 8 (Muthen & Muthen, 1994). Maximum likelihood (ML) parameter estimation was chosen over other estimation methods (weighted least squares, two-stage least squares) because the data were distributed normally, and ML estimation gives unbiased parameter estimates and standard errors for any missing data in the dataset (Schreiber, Nora, Stage, Barlow, & King, 2006). SEM estimated data using full maximum likelihood parameter estimation for 169 missing data points on the measure of maternal lifestyle behaviors (HPLP-II).


23
Three models were tested using Mplus to test three hypotheses. The first model tested the first hypothesis of the theoretical dimensional structure of a latent construct, named “socioeconomic resources”, and whether four constructs of socioeconomic resource—personal income, family income, subjective economic hardship measured by the Economic Hardship Scale (EHS; Barrera et al.,) and subjective social status measured by the McArthur Scale of Subjective Social Status (SSS; Adler & Stewart, 2007)—significantly comprise the underlying factor structure. The first model also tested the second hypothesis that the socioeconomic resources latent factor predicts engagement in maternal health behaviors, measured by the Health Promotion Lifestyle Profile (HPLP; Walker, Noble, Sechrist, Richert, & Pender, 1987).
In the second and third models, the third hypothesis was tested and used SEM to explore the association of proposed mediators on the relationship between socioeconomic resources and maternal engagement in lifestyle behaviors. In the second model SEM was conducted to test whether a proposed mediator, maternal depression, assessed using the Edinburgh Postnatal Depression Scale (EPDS; Cox et al., 1987), mediates the relationship between the socioeconomic resources latent factor and engagement in maternal health behaviors. Lastly, in the third model, SEM was conducted to test a post hoc hypothesis of whether two binary variables, years of education (high: 12 years or more of education; low: less than 12 years of education) and race (White; non-White), significantly mediate the relationship between socioeconomic resources and maternal lifestyle behaviors.


24
CHAPTER THREE Results
Sample Characteristics
All participants had at least one child at or under the age of 10 years (exclusionary criteria if they did not). More than half of the sample (55.7%) were between the ages of 25 and 34 years. Thirty four percent of the sample were between the ages of 35 and 44, 8.9% of the sample were between the ages of 15 and 24 and 1.3% of the sample were between the ages of 45 and 55. Eighty four percent of mothers were of White or Caucasian decent, 11% of mothers were of Hispanic or Latino decent, less than 1% were of African American or Black decent, less than 2% were of Asian American or Pacific Islander decent, about 3% were of multiracial decent and none were of American Indian or Alaskan Native decent. The majority of our sample had 12 or more years of education (83%), and were employed (80.5%). 56.6% of the sample had a personal income level of $39,999 or less while 43.4% had a personal income of $40,000 or more. Additionally, 47.5% of the sample had a combined family income of $79,000 or less and 52.5% of the sample had a combined family income of $80,000 or more. The majority of the sample were either married or in a significant romantic relationship (89.9%).
From the sample of 159 mothers, 102 mothers completed the entire survey, 46 did not fill out one or more full scales and 11 completed the survey with the exception of a few single missing times. For participants that were missing single items, data was imputed based on the average of other items in the same scale. Only one of the scales, the composite score for the Economic Hardship Scale (EHS) had a non-normal distribution, with a skewness of 1.26 and kurtosis of 1.01.
Association among Study Variables


25
Primary Analysis
Model 1
The first model of the study examined the influence of economic hardship on maternal health-promoting lifestyle behaviors. Model 1, using internal health locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, is positive but not statistically significant, c’ = .01, t (107) = .31, p = .755. Table 4 and figure 2 show significant indirect effects found in this model include the relationships between: 1) economic hardship, internal health locus of control, health motivation and lifestyle behaviors was significant (c ’cs denotes a completely standardized effect size, c cs = -.01; Hayes, 2013), 2) economic hardship, internal health locus of control, health self-efficacy and lifestyle behaviors was also significant (c ’cs = -.01), 3) economic hardship, internal health locus of control, task oriented coping and lifestyle behaviors was also significant (c’cs =-.01) and lastly, 4) economic hardship, internal health locus of control, health motivation, health self-efficacy and lifestyle behaviors was significant with an effect size of -.002. Secondly, table 12 and figure 10 shows the direct effect of a separate analysis in model 1, that used maternal depression and learned helplessness as mediators, was also positive but not statistically significant, c’ = .01, t (107) = .312 p = .755; there were no significant indirect effects found.
Model 2
The second model of the study measured socioeconomic status, both personal and familial, and educational achievement on maternal health-promoting lifestyle behaviors. A separate analysis in model 2 also measured individual perception of social status to further understand another dimension of perception of hardship. Table 5 and figure 3 show the direct effect of model 2, with educational achievement as the independent variable, and internal health


26
locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, was positive but not statistically significant, c’ = -.06, t (107) = -1.37, p = .176. The indirect effects of the relationship between years of education, health motivation, health self-efficacy, task oriented coping and lifestyle behaviors {c’cs = .001), as well as years of education, task oriented coping and lifestyle behaviors were significant (c ’cs = .06). Table 13 and figure 11 show the direct effect of a separate analysis in model two, using maternal depression and learned helplessness as mediators, was negative but not statistically significant, c’ = -.06, t (107) = -1.37, p = .176.
Table 6 and figure 4 show the direct effect of model 2, with annual personal income as the independent variable, and internal health locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, was positive but not statistically significant, c’ = .04, t (107) = 1.96, p = .053; significant indirect effects included the relationships between 1) personal income, internal health locus of control, health motivation and lifestyle behaviors (c ’cs = .01), 2) personal income, internal locus of control, health self-efficacy and lifestyle behaviors (c ’cs = .01), 3) personal income, internal health locus of control, task oriented coping, and lifestyle behaviors {c’cs = .01), 4) personal income, internal health locus of control, health motivation, health self-efficacy and lifestyle behaviors (c’cs = .003), 5) personal income, internal health locus of control, health self-efficacy, task oriented coping, and lifestyle behaviors (c ’cs = .0004), and lastly, 6) personal income, health motivation, health self-efficacy, task oriented coping and lifestyle behaviors (c’cs= .0004).Table 14 and figure 12 show the direct effect of model 2, with annual personal income as the independent variable, and maternal depression and learned helplessness as mediators, was positive but not statistically significant c’ = .04, t (107) =
1.96, p = .053.


27
Table 7 and figure 5 show the direct effect of model 2 with annual family income as the independent variable, and internal health locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, was positive but not statistically significant, c’ = .02, t (107) = .87, p = .385. The indirect effect of the relationship between family income, health motivation, health self-efficacy, task oriented coping and lifestyle behaviors was significant (c ’cs = .001). Table 15 and figure 13 show the direct effect of model 2 with annual family income as the independent variable, and maternal depression and learned helplessness as mediators, was positive but not statistically significant, c’ = .02, t (107) = .87, p = .385; there were no significant indirect effects.
Table 8 and figure 6 show the direct effect of model 2 with social support as the independent variable, and internal health locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, was positive but not statistically significant, c’ = .02, t (107) = .63, p = .530. The indirect effects of the relationship between social support, internal locus of control, health self-efficacy, task oriented coping and lifestyle behaviors (c ’cs = .001) and social support, internal health locus of control, health motivation, health self-efficacy, task oriented coping and lifestyle behaviors (c ’cs = .0002) were statistically significant. Table 16 and figure 14 show the direct effect of model 2 with social support as the independent variable, and maternal depression and learned helplessness as mediators, was also positive but not statistically significant c’ = .02, t (107) = .62, p = .530; there were no significant indirect effects.
Table 9 and figure 7 show the direct effect of model 2 with subjective social status, current, as the independent variable, and internal health locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, was negative but not statistically significant, c’ = -.01, t (107) = -.56, p = .577; there were no significant indirect effects. Table 17


28
and figure 15 show the direct effect of model 2 with subjective social status, current, as the independent variable, and maternal depression and learned helplessness as mediators, was also negative but not statistically significant c’ = -.01, t (107) = -.56, p = .577 with no significant direct effects.
Table 10 and figure 8 show the direct effect of model 2 with subjective social status, future, as the independent variable, and internal health locus of control, health motivation, health self-efficacy and task-oriented coping as mediators, was positive but not statistically significant, c’ = .01, t (107) = .23, p = .823. There were four indirect effects found to be significant in this model: 1) the indirect effect between subjective social status, future, internal health locus of control, health motivation, health self-efficacy and lifestyle behaviors was significant (c ’cs = .002) 2) indirect effect between subjective social status, future, internal health locus of control, health self-efficacy, task-oriented coping and lifestyle behaviors was significant (c ’cs = .001); 3) the indirect effect between subjective social status, internal health locus of control, health motivation, heath self-efficacy, task-oriented coping and healthy lifestyle behaviors was significant (c ’cs = .0003); 4) the indirect effect between subjective social status, future, health motivation, health self-efficacy, task-oriented coping and lifestyle behaviors was significant (c ’cs = .001). Table 18 and figure 16 show the direct effect of model 2 with subjective social status, future, as the independent variable, and maternal depression and learned helplessness as mediators, was positive but not statistically significant c’ = .01, t (107) = .23, p = .823.
Model 3
Thirdly, access to health resources was a third model of the study. Table 11 and figure 9 show the direct effect of model 3 with access to health resources as the independent variable, and internal health locus of control, task-oriented coping, health self-efficacy and health motivation


29
as mediators, was positive but not statistically significant, c’ = .04, t (107) = .96, p = .338. There were six significant indirect effects found in this model: 1) access to healthcare resources, internal health locus of control, health motivation and lifestyle behaviors (c ’cs = .02) 2) the relationship between access to healthcare resources, internal health locus of control, health self-efficacy and lifestyle behaviors (c’cs = .01), 3) the relationship between access to healthcare resources, internal health locus of control, task oriented coping, and lifestyle behaviors (c ’cs = .01), 4) the relationship between access to healthcare resources, internal health locus of control, health motivation, health self-efficacy and lifestyle behaviors (c ’cs = .003), 5) the relationship between access to healthcare resources, internal health locus of control, health self-efficacy, task oriented coping and lifestyle behaviors (c ’cs = .002), 6) the relationship between access to healthcare resources, internal health locus of control, health motivation, health self-efficacy, task oriented coping and lifestyle behaviors (.001). Table 19 and figure 17 show the direct effect of model 3 with access to health resources as the independent variable, and maternal depression and learned helplessness as mediators, was also positive but not statistically significant, c’ = .04, t (107) = .96, p = .338 with no significant indirect effects.
Secondary Analysis
We ran a structural equation model to test the hypotheses that 1) a socioeconomic resources variable would have good factor structure using personal income, family income, subjective economic hardship and subjective social status, 2) that the socioeconomic resources latent variable would significantly predict maternal engagement in health promoting lifestyle behaviors and 3) that maternal depression, years of education and race would also significantly mediate this relationship. First, fit statistics of the first model (figure 22) confirmed that the data fit the model for each component of socioeconomic resources and the prediction of the latent


30
construct of socioeconomic resources on maternal engagement in health promoting lifestyle behaviors, Chi Squared (5, N = 288) = 10.76, p = .056, root mean square error of approximation (RMSEA) = .06, comparative fit index (CFI) = .97. Fit statistics of the second model (figure 23) confirmed adequate fit of maternal depression mediating the relationship between the latent construct of socioeconomic resources and maternal engagement in health promoting lifestyle behaviors, Chi Squared (8, N = 288) = 22.44, p = .004, RMSEA = .08, CFI = .94, with a nonsignificant standardized indirect effect of -.01, p = .423. Fit statistics for the third model (figure 24) confirmed good fit of years of education and race mediating the relationship between the latent construct of socioeconomic resources and maternal engagement in health promoting lifestyle behaviors, Chi Squared (12, 288) = 17.53, p = .131, RMSEA = .04, CFI = .98, with no indirect effects reported in the model due to the binary variables used.
Post Hoc Analyses
Economic Hardship Subscale scores
Post hoc analyses were conducted using the four subsets of the economic hardship scale as the independent variables instead of the economic hardship composite score that was used in the main analyses. Table 20 and figure 18 show the direct effect of the post hoc analysis with inability to make ends meet as the independent variable and internal health locus of control, task-oriented coping, health self-efficacy and health motivation as mediators, was positive but not statistically significant, c’ = .09, t (107) = 1.13, p = .263 with no significant indirect effects.
Table 21 and figure 19 show the direct effect of the post hoc analysis with adjustments and cutbacks as the independent variable and internal health locus of control, task-oriented coping, health self-efficacy and health motivation as mediators, was also positive but not statistically significant, c’ = .04, t (107) = 1.22, p = .227. Table 22 and figure 20 show the direct effect of the


31
post hoc analysis with not enough money for necessities as the independent variable and internal health locus of control, task-oriented coping, health self-efficacy and health motivation as mediators, was negative and statistically significant, c’ = -.22, t (107) = -.2.84, p = .006. Finally, table 20 and figure 20 show the direct effect of the post hoc analysis with financial strain as the independent variable and internal health locus of control, task-oriented coping, health self-efficacy and health motivation as mediators, was positive but not statistically significant, c’ =
.02, t (107) = . 19, p = .847.


32
CHAPTER FOUR Discussion
The purpose of the current study was to examine the influence of psychological variables on the relationship between demographic and socioeconomic variables and maternal lifestyle behaviors. As a means to further understand the proximate causes that create health disparities, the current study hypothesized that psychological mediators would significantly influence the relationship between sociodemographic and socioeconomic factors and maternal lifestyle behaviors. I did not find any significant direct mediation effects of psychological variables on the relationship between sociodemographic and socioeconomic variables and maternal lifestyle behaviors in the current sample. The current study identified three separate models to: 1) examine the influence of a subjective perspective of economic hardship on lifestyle behaviors through the influence of psychological mediators, 2) assess the relationship between socioeconomic status, educational achievement and social support on maternal lifestyle behaviors through the influence of psychological mediators and 3) assess the relationship between maternal access to health resources, through the influence of psychological mediators, on maternal health behaviors. The results showed that despite insignificant findings of the overall mediating effects of psychological variables on the relationship between specific demographic and socioeconomic factors and maternal lifestyle behaviors, there are statistically significant indirect pathways between the variables that were investigated.
The results of the current study show tentative indirect relationships between specific psychological mediators and maternal lifestyle behaviors. Relative to those with lower subjective experience of hardship, those with a higher subjective perspective of hardship had a lower engagement in health promoting lifestyle behaviors through the pathways of lower internal


33
health locus of control, lower health motivation, health self-efficacy and task oriented coping style. Higher educational achievement resulted significant indirect pathways of higher health motivation, health self-efficacy, and task oriented coping, leading to higher engagement in health promoting lifestyle behaviors. Additionally, those with higher annual personal income had higher engagement in health promoting lifestyle behaviors through the pathways of internal health locus of control, health motivation, health self-efficacy and task oriented coping. Family income also had one significant indirect effect showing higher family income resulting in higher engagement in health promoting lifestyle behaviors through the pathways of higher health motivation, task oriented coping, and health self-efficacy. Social support played a similar role, with higher social support leading to higher engagement in health promoting lifestyle behaviors through the pathways of higher internal health locus of control, health motivation, health self-efficacy, and task oriented coping. An additional measure of subjective perception of hardship, anticipated future subjective social status, was significantly indirectly and positively related to maternal health behaviors through multiple pathways, including internal health locus of control, task-oriented coping, health motivation, and heath self-efficacy. Lastly, higher access to health resources also resulted in higher engagement in health promoting lifestyle behaviors through pathways of higher internal health locus of control, health motivation, health self-efficacy and task oriented coping.
Results of three models of SEM in the secondary analysis conclude that 1) the underlying factor structure of the latent construct fit the data well and the latent construct of socioeconomic resources significantly predicts maternal engagement in lifestyle behaviors, 2) maternal depression does not significantly mediate the relationship between the latent construct socioeconomic resources and maternal engagement in health promoting lifestyle behaviors and


34
3) years of education and race do not significantly mediate the relationship between the latent construct socioeconomic resources and maternal engagement in health promoting lifestyle behaviors.
Post hoc analyses also found that one subset of the EHS significantly predicted engagement in health promoting lifestyle behaviors. The not enough money for necessities subset score significantly negatively predicted engagement in health promoting lifestyle behaviors, showing this subset of the EHS may capture a monetary predictor of engagement in lifestyle behaviors. Questions such as “We had enough money to afford the kind of food we should have” and “We had enough money to afford leisure and recreational activities” may capture a better depiction of monetary resources that result in an increase or decrease in engagement in healthy lifestyle behaviors. However, this model displays a significant direct effect of not enough money for necessities and maternal engagement in health-promoting lifestyle behaviors; a full mediation is not supported since the independent variable does not significantly predict the psychological mediators.
One potential explanation for why the current study did not find significant direct main effects may be due to the demographics of the sample. The recruitment method was unable capture a diverse sample of mothers from both high and low socioeconomic backgrounds, ethnically and racially diverse backgrounds and of a diverse group of educational and occupational experiences. Future research may examine the direct relationships between the demographic and socioeconomic variables and psychological mediators, and maternal lifestyle behaviors. It may be that the various demographic and socioeconomic variables significantly predict maternal lifestyle behaviors independently, as shown in previous research (Matijasevich et al., 2010; Premji, 2014). Psychological mediators may also independently predict maternal


35
engagement in lifestyle behaviors. It may be helpful to investigate the relationship between psychological mediators and maternal lifestyle behaviors while controlling for demographic and socioeconomic variables. Future research should primarily establish these two relationships prior to conducting a mediational analysis combining the target variables.
To date, the current study is the first to look at multiple mediating variables that may influence the relationship between sociodemographic and socioeconomic risk factors on lifestyle behaviors in a sample of mothers. Although direct model findings were insignificant, the current study tentatively provides information on indirect pathways through which psychological mediators may influence maternal lifestyle behaviors.
Implications
The purpose of the current study was to examine the relationships between demographic and socioeconomic factors and maternal engagement in health-promoting lifestyle behaviors. Furthermore, the current study aimed to investigate whether individual attitudes, or specific psychological variables, influence the above relationship.
The current study did not provide support for the HSE theory in understanding health-promoting lifestyle behaviors and levels of engagement. Specifically, the aim of the study was to extend the findings of previous research that shows these self-empowerment or self-motivation behaviors are particularly influential on lifestyle behaviors and choices. Furthermore, our mediation hypothesis was not supported. However, the significant indirect effects found may help to identify tentative mechanistic pathways that explain why some women engage in health-promoting lifestyle behaviors despite dealing with socioeconomic and sociodemographic stressors. Previous research in self-motivation attitudes has found that specific factors may influence attitudes. Chang et al. (2011) found that availability of time mothers had to prepare


36
food affected their engagement in health-promoting dietary behaviors, and a higher positive mood increased self-efficacy and healthy choices. Additionally, maternal outcome of expectancies may also affect maternal engagement in healthy lifestyle choices. Chang et al. (2008) also found that if mothers are aware of the beneficial effects of healthy choices, they are more likely to engage in health-promoting behaviors despite dealing with challenges associated with low-income or underserved backgrounds. Social support may also impact attitudes toward behaviors and the success of behavioral interventions, as Clarke et al. (2007) found positive social support to be a significant contributor to engagement in healthy lifestyle choices and behavior, specifically in mothers.
The findings of the current study suggest that variables such as self-efficacy, health motivation, task-oriented coping and locus of control are predictive of maternal health behaviors (b pathways), but do not significantly influence the relationship between the predictor and outcome variables. Furthermore, it may be that the proposed mediators serve as independent variables and covariates along with the other proposed independent variables in the current study. This is in line with the previous literature mentioned above, as well as a possible explanation of the null findings. Instead of serving as the mechanisms through which demographic and socioeconomic factors influence maternal lifestyle behaviors, it may make more sense to view psychological mediators as influences on maternal lifestyle behaviors on their own.
If future studies are able to replicate this model with a population of specifically low-income mothers, it may be that the proposed mediators do influence the relationship between demographic and socioeconomic factors and maternal lifestyle behaviors. Through the increased engagement in health-promoting lifestyle behaviors, the current study was able to show how to


37
potentially maximize health-promoting lifestyle behaviors in mothers, through the influence of psychological mediators. This may be able to help healthcare providers, social workers, psychologists, teachers and other services workers within low-income communities to assess for and fosters health-promotion focused lifestyle choices and behaviors through the use of these psychological mediators. Teaching skills in health motivation and health self-efficacy, for example, through health behavior interventions, may potentially prevent serious and chronic health problems, such as cardiovascular disease, diabetes, and a variety of psychological disorders, in a variety of populations by increasing engagement of health promoting lifestyle behaviors (Walsh, 2011). Modeling of maternal engagement in health-promoting lifestyle behaviors in low-income and underserved communities may also have even greater implications in health outcomes by teaching children and younger generations about health-promoting lifestyle behaviors and choices which may eventually decrease health disparities in this population.
Study Limitations
The findings of the current study need to be interpreted with caution based on some limitations of the study. First, the current study’s aim was to recruit a diverse sample of mothers from underserved backgrounds. Although successful in recruiting a diverse group of mothers in some demographics, such as age, recruitment via Facebook was not successful in reaching mothers from significantly underserved backgrounds (e.g., annual income of <$10,000). Additionally, based on the demographic results of the analysis, our sample primarily consisted of Caucasian women (83.6%), therefore, external validity is threatened. The mechanistic pathways that were conditionally found in a somewhat homogeneous sample may be more clearly defined in a more diverse sample that includes more women from lower income backgrounds. The use of


38
Facebook for recruitment threatened external validity and generalizability of results and created a selection bias that affected the independent variables by the voluntary and self-selected recruitment method (e.g., demographic and socioeconomic factors and the mediators, self-efficacy, locus of control, etc.). Additionally, convenience sample from Facebook recruitment skewed the data in that we were unable to capture women who do not have Facebook or access to technology and social media.
Thirdly, the use of a cross-sectional research design is a significant limitation of the study. An important point of mediation analyses is that the independent variable must precede the dependent variable, which inherently implies causation, with mediator variables thought to be causal variants between the criterion and outcome variables. Previous research has investigated whether mediation can be tested with cross-sectional data. Some researchers have concluded that the results of analyses based on cross-sectional data are unlikely to accurately reflect longitudinal mediation effects, and that using longitudinal designs enables researchers to make more rigorous inferences about causal relations implied by mediation models (Cole & Maxwell, 2003). The results of the current study may reflect a bias or misleading estimate of the relationship between the target variables. If socioeconomic and demographic variables, psychological mediators and maternal lifestyle behaviors had been studied at two or three different time points, different mediation results may have been shown.
Potential threats to the internal and statistical conclusion validity of the study also deserve mention. Corrections of controlling for inflation and type I error were planned in the analysis and methods sections by splitting up the independent variables into three separate analyses and using a Hayes multiple mediation analysis to decrease the number of regressions run and minimize increased inflation of type one and two errors for each analysis being done. A


39
bias-corrected confidence interval of 90% was also used. However, it is still possible that we failed to detect significant results because the current study did not have adequate power with the current sample size relative to the number of variables used. Additionally, because the scales being used in the online survey have been formally used and validated in paper format, this factor may also pose a threat to internal validity. Psychometric tests (e.g., assessing for Cronbach’s alpha, replicating factor structures in online forms with paper forms) to assess validity and reliability were performed during analysis to assess for this potential threat to validity. One scale was particularly low in reliability. The Cronbach’s alpha for our access to health resources items was .59, indicating a significantly lower reliability between the scale items. However, although there were some consequences of changing the mode of administration for the surveys being used (e.g., difficult to control study environment, unmonitored study participants, and selection bias), there have been found to be benefits of online research studies. Greater population access, less experimental costs, no time limitations and complete voluntary participation which may indicate participant motivation (Musch & Reips, 2000) are examples of beneficial differences that were found to make internet-based data collection a suitable alternative to in-person administration, keeping in mind the data collection medium creates a suitable but not identical alternative (Riva, Teruzzi, & Anolli, 2003).


40
REFERENCES
Abel, M. H. (1996). Inadequate Prenatal Care. Psychological Reports, 79, 903-912.
Adler, N., & Stewart, J. (2007). The MacArthur scale of subjective social status. MacArthur Research Network on SES & Health. Retrieved from http: innr. macses. ucsf. eduResearch Psychosocial/subjective, php.
Abusabha, R., & Achterberg, C. (1997). Review of self-efficacy and locus of control for nutrition-and health-related behavior. Journal of the American Dietetic Association, 97(10), 1122-1132.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-Hall, Inc.
Baron, R.M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of personality and social psychology, 51(6), 1173.
Barrera, M., Caples, H., and Tein, J. (2001). The psychological sense of economic hardship: Measurement models, validity, and cross-ethnic equivalence for urban families. American Journal of Community Psychology, 29(3), 493-517.
Bergink, V., Kooistra, L., Lambregtse-van den Berg, M. P., Wijnen, H., Bunevicius, R., van Baar, A., & Pop, V. (2011). Validation of the Edinburgh Depression Scale during pregnancy. Journal of psychosomatic research,70(A), 385-389.
Biddle, S. J., & Asare, M. (2011). Physical activity and mental health in children and
adolescents: a review of reviews. British Journal of Sports Medicine, -/5(11), 886-95. http://doi.org/10.1136/bjsports-2011-090185
Blair, C., Granger, D. A., Willoughby, M., Mills-Koonce, R., Cox, M., Greenberg, M. T., ... Fortunato, C. K. (2011). Salivary Cortisol Mediates Effects of Poverty and Parenting on Executive Functions in Early Childhood. Child Development, 82(6), 1970-1984. http://doi.org/10.1111/j. 1467-8624.2011.01643.x
Buford, T. W., Anton, S. D., Clark, D. J., Higgins, T. J., & Cooke, M. B. (2014). Optimizing the benefits of exercise on physical function in older adults. PM & R : The Journal of Injury, Function, and Rehabilitation, 6(6), 528-43. http://doi.Org/10.1016/j.pmrj.2013.ll.009
Chang, M. W., Nitzke, S., Brown, R., & Baumann, L. (2011). Predictors of Low-income, Obese Mothers’ Use of Healthful Weight Management Behaviors. Journal of Nutrition Education and Behavior, 43(2), 87-95. http://doi.org/10.1016/jjneb.2009.ll.006
Clarke, K. K., Freeland-Graves, J., Klohe-Lehman, D. M., & Bohman, T. M. (2007). Predictors of Weight Loss in Low-Income Mothers of Young Children. Journal of the American Dietetic Association, 107(1), 1146-1154. http://doi.org/10.1016/jjada.2007.04.016


41
Clearfield, M. W., Carter-Rodriguez, A., Merali, A. R., & Shober, R. (2014). The effects of SES on infant and maternal diurnal salivary cortisol output. Infant Behavior and Development, 57(3), 298-304. http://doi.Org/10.1016/i.infbeh.2014.04.008
Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: questions and tips in the use of structural equation modeling. Journal of abnormal psychology, 772(4), 558.
Corwin. (2008). Immune dysregulation and glucocorticoid resistance in minority and low-income pregnancy women. October 141(4), 520-529. http://doi.Org/10.1016/j.surg.2006.10.010.Use
Costa, F. M., lessor, R., & Donovan, J. E. (1989). Value on health and adolescent
conventionality: A construct validation of a new measure in problem-behavior theory. Journal of Applied Social Psychology, 79(10), 841-861. http ://doi. org/10.1111/j. 1559-1816.1989.tb01225.x
Cosway, R., Endler, N. S., Sadler, A. J., & Deary, I. J. (2000). The coping inventory for stressful situations: Factorial structure and associations with personality traits and psychological health1. Journal of Applied Bio-Behavioral Research, 5(2), 121-143.
Cox, J. L., Holden, J. M., & Sagovsky, R. (1987). Detection of postnatal depression.
Development of the 10-item Edinburgh Postnatal Depression Scale. The British journal of psychiatry, 150(6), 782-786.
Dayan, J., Creveuil, C., Marks, M. N., Conroy, S., Herlicoviez, M., Dreyfus, M., & Tordjman, S. (2006). Prenatal depression, prenatal anxiety, and spontaneous preterm birth: a prospective cohort study among women with early and regular care. Psychosomatic Medicine, 68(6), 938-946.
Dennis, J. (2013). Young maternal age and low birth weight risk: An exploration of racial/ethnic disparities in the birth outcomes of mothers in the United States, 141(4), 520-529. http://doi.Org/10.1016/j.surg.2006.10.010.
Dog, L. (2010). The role of nutrition in mental health. Alternative therapies in health and medicine, 16(2), 42.
Dole, N., Savitz, D. A., Herz-Picciotto, I., Siega-Riz, A. M., McMahon, M. J., & Buekens, P. (2003). Maternal Stress and Preterm Birth. American Journal of Epidemiology, 157(1), 14-24. http://doi.org/10.1093/aje/kwfl76
Dominguez, T. P. (2011). Adverse birth outcomes in African American women: the social
context of persistent reproductive disadvantage. Social Work in Public Health, 26(1), 3-16. http://doi.org/10.1080/10911350902986880


42
Duffy, M. E. (1997). Determinants of reported health promotion behaviors in employed Mexican American women. Health Care for Women International, 18(2), 149-163. http://doi.org/10.1080/07399339709516270
Eberhard-Gran, M., Eskild, A., Tambs, K., Opjordsmoen, S., & Ove Samuelsen, S. (2001). Review of validation studies of the Edinburgh Postnatal Depression Scale. Acta Psychiatrica Scandinavica, 104(4), 243-249.
Essex, M., Klein, M., Cho, E., & Kalin, N. H. (2002). Maternal stress beginning in infancy may sensitize children to later stress exposure: Effects on cortisol and behavior. Biological Psychiatry, 52(8), 776-784.
Evans, G. (2003). A multimethodological analysis of cumulative risk and allostatic load among rural children. Developmental Psychology, 39(5), 924-933. http://dx.doi.org/10.1037/0012-1649.39.5.924
Everhart, K., & Emde, R. N. (2006). Perspectives on stress and self-regulatory processes. The Crisis in Youth Mental Health: Childhood disorders, 1, 1.
Fink, N. S., Urech, C., Cavelti, M., & Alder, J. (2012). Relaxation During Pregnancy. The Journal of Perinatal & Neonatal Nursing, 26(4), 296-306. http://doi.org/10.1097/JPN.0b013e31823f565b
Fritz, M. S., & MacKinnon, D. P. (2007). Required sample size to detect the mediated effect. Psychological science, 18(f), 233-239.
Froman, R. D., & Owen, S. V. (1991). High school students' perceived self-efficacy in physical and mental health. Journal of Adolescent Research, 6(2), 181-196.
Fiscella, K., & Williams, D. R. (2004). Health disparities based on socioeconomic inequities: implications for urban health care. Academic Medicine: Journal of the Association of American Medical Colleges, 79(12), 1139-1147. http://doi.org/10.1097/00001888-200412000-00004
Gazmararian, J. A., Adams, M. M., & Pamuk, E. R. (1996). Associations between measures of socioeconomic status and maternal health behavior. American Journal of Preventive Medicine, 12(2), 108-15. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8777063
Giatti, L., Camelo, L. D. V., Rodrigues, J. F. D. C., & Barreto, S. M. (2012). Reliability of the MacArthur scale of subjective social status - Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). BMC Public Health, 72(1), 1096. http://doi.org/10.1186/1471-2458-12-1096
Gennaro, S. (2003). Psychological and Physiological Stress: Impact on Preterm Birth. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 32(5), 668-675. http://doi.org/10.1177/0884217503257484


43
Ghosh, J. K. C., Wilhelm, M. H., Dunkel-Schetter, C., Lombardi, C. A., & Ritz, B. R. (2010). Paternal support and preterm birth, and the moderation of effects of chronic stress: A study in Los Angeles County mothers. Archives of Women’s Mental Health, 13(4), 327-338. http ://doi. org/10.1007/s0073 7-009-0135-9
Gortmaker, S. L. (1979). The effects of prenatal care upon the health of the newborn. American Journal of Public Health, 69(1), 653-660. http://doi.Org/10.2105/AJPH.69.7.653
Grossman, P., Niemann, L., Schmidt, S., & Walach, H. (2004). Mindfulness-based stress
reduction and health benefits: A meta-analysis. Journal of Psychosomatic Research, 57(1),
35-43. http://doi.org/10.1016/S0022-3999(03)00573-7
Guedeney, N., & Fermanian, J. (1998). Validation study of the French version of the Edinburgh Postnatal Depression Scale (EPDS): new results about use and psychometric properties. European psychiatry, 13(2), 83-89.
Haller, D. L., Knisely, J. S., Dawson, K. S., & Schnoll, S. H. (1993). Perinatal substance abusers: Psychological and social characteristics. The Journal of Nervous and Mental Disease, 757(8), 509-513.
Hardie, J. H., & Landale, N. S. (2013). Profiles of Risk: Maternal Health, Socioeconomic Status, and Child Health. Journal of Marriage and Family, 75(3), 651-666. http ://doi. org /10.1111 /j omf. 12021
Hartnett, H. &. (2011). Who lacks support and why? An examination of mothers’ personal safety nets. Journal of Marriage and Family, 73(4), 861-875. http://doi.Org/10.1038/naturel3314.A
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press.
Hein, A., Rauh, C., Engel, A., Haberle, L., Dammer, U., Voigt, F., ... Goecke, T. W. (2014). Socioeconomic status and depression during and after pregnancy in the Franconian Maternal Health Evaluation Studies (FRAMES). Archives of Gynecology and Obstetrics, 289(4), 755-63. http://doi.org/10.1007/s00404-013-3046-y
Hiroto, D. S. & Seligman, M.E. (1975). Generality of learned helplessness in man. Journal of Personality and Social Psychology, Vol 31(2), 311-327. http://dx.doi.org/10.1037/h007627Q
Hobel, C. J., Dunkel-Schetter, C., Roesch, S. C., Castro, L. C., & Arora, C. P. (1999). Maternal plasma corticotrophin-releasing hormone associated with stress at 20 weeks’ gestation in pregnancies ending in preterm delivery. American journal of obstetrics and gynecology, 750(1), S257-S263.


44
Jadresic, E., Araya, R., & Jara, C. (1995). Validation of the Edinburgh postnatal depression scale (EPDS) in Chilean postpartum women. Journal of Psychosomatic Obstetrics & Gynecology, 16(A), 187-191.
James, S. A. (1993). Racial and ethnic differences in infant mortality and low birth weight. A psychosocial critique. Annals of Epidemiology, 3, 130-136. http ://doi. org/http ://dx.doi. org/10.1016/1047-2797(93 )90125-N
Jayanti, R. K., & Burns, A. C. (1998). The Antecedents of Preventive Health Care Behavior: An Empirical Study. Journal of the Academy of Marketing Science, 26(1), 6-15. http ://doi. org/10.1177/0092070398261002
Jordan, K. C., Freeland-Graves, J. H., Klohe-Lehman, D. M., Cai, G., Voruganti, V. S., Proffitt, J. M., ... Bohman, T. M. (2008). A nutrition and physical activity intervention promotes weight loss and enhances diet attitudes in low-income mothers of young children. Nutrition Research, 28(1), 13-20. http://doi.Org/10.1016/j.nutres.2007.ll.005
Kahn, H. S., Williamson, D. F., & Stevens, J. A. (1991). Race and weight change in US women: The roles of socioeconomic and marital status. American Journal of Public Health, 81(3), 319-323. http://doi.org/10.2105/AJPH.8T3.319
Lancaster, C. A., Gold, K. J., Flynn, H. A., Yoo, H., Marcus, S. M., & Davis, M. M. (2010). Systematic Review, 202(1), 5-14. http://doi.Org/10.1016/j.ajog.2009.09.007.Risk
Lantz, Paula M., et al. "Socioeconomic disparities in health change in a longitudinal study of US adults: the role of health-risk behaviors." Social science & medicine 53.1 (2001): 29-40.
Luecken, L. J. (2008). Prenatal Stress, Partner Support, and Infant Cortisol Reactivity in Low-Income Mexican American Families. October 141(A), 520-529. http://doi.Org/10.1016/i.surg.2006.10.010.
Luecken, L. J., Lin, B., Coburn, S. S., MacKinnon, D. P., Gonzales, N. A., & Crnic, K. A.
(2013). Prenatal stress, partner support, and infant cortisol reactivity in low-income Mexican American families. Psychoneuroendocrinology, 55(12), 3092-3101.
Mailey, E. L., & McAuley, E. (2014). Impact of a brief intervention on physical activity and social cognitive determinants among working mothers: A randomized trial. Journal of Behavioral Medicine, 37(2), 343-355. http://doi.org/10.1007/sl0865-013-9492-y
Matijasevich, A., Victora, C. G., Lawlor, D. A., Golding, J., Menezes, A. M. B., Araujo, C. L., Barros, A. J., Santos, I. S., Barros, F. C., & Smith, G. D. (2012). Association of socioeconomic position with maternal pregnancy and infant health outcomes in birth cohort studies from Brazil and the UK. Journal of Epidemiology and Community Health, 66(2), 127-35. http://doi.org/10.1136/jech.2010.108605


45
Matthews, K. A., Gallo, L. C., & Taylor, S. E. (2010). Are psychosocial factors mediators of socioeconomic status and health connections? Annals of the New York academy of sciences, 7756(1), 146-173.
McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiological Reviews, 87, 873-904. http://doi.org/10.1152/physrev.00041.2006
Miller, G. E., Chen, E., Fok, A. K., Walker, H., Lim, A., Nicholls, E. F., ... Kobor, M. S. (2009). Low early-life social class leaves a biological residue manifested by decreased glucocorticoid and increased proinflammatory signaling. Proceedings of the National Academy of Sciences of the United States of America, 706(34), 14716-21. http://doi.org/10.1073/pnas.0902971106
Minkovitz, C. S., O’Campo, P. J., Chen, Y.-FL, & Grason, H. A. (2002). Associations between maternal and child health status and patterns of medical care use. Ambulatory Pediatrics, 2(2), 85-92. http://doi.org/10.1367/1539-4409(2002)002<0085:abmach>2.0.co;2
Mitchell, T. M. D. F., & Barlow, C. E. M. S. (2011). Review of the Role of Exercise in Improving Quality of Life in Healthy Individuals and in Those with Chronic Diseases. Current Sports Medicine Reports, 70(4), 211-216. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=ovftl &AN=00149619-201107000-00011
Musch, J., & Reips, U. D. (2000). A brief history of Web experimenting. Retrieved from psychnet.apa.org
Parker, J. D., Schoendorf, K. C., & Kiely, J. L. (1994). Associations between measures of socioeconomic status and low birth weight, small for gestational age, and premature delivery in the United States. Annals of Epidemiology, -7(4), 271-278. http ://doi. org/10.1016/1047-2797(94)90082-5
Prather, H., Spitznagle, T., & Hunt, D. (2012). Benefits of exercise during pregnancy. PM and R, -7(11), 845-850. http://doi.Org/10.1016/j.pmrj.2012.07.012
Premji, S. (2014). Perinatal Distress in Women in Low- and Middle-Income Countries:
Allostatic Load as a Framework to Examine the Effect of Perinatal Distress on Preterm Birth and Infant Health. Maternal and Child Health Journal, 75(10), 2393-2407. http ://doi. org/10.1007/s 10995-014-1479-y
Pop, V. J., Komproe, I. H., & Van Son, M. J. (1992). Characteristics of the Edinburgh post-natal depression scale in The Netherlands. Journal of affective disorders, 26(2), 105-110.
Program for prevention research. (1999). Manual for the children’s coping strategies checklist and the how I coped under pressure scales. Available from Arizona State University, P.O. Box 876005, Tempe, AZ 85287-6005, 7720(480).


46
Quinless, F. W., & Nelson, M. A. M. (1988). Development of a measure of learned helplessness. Nursing Research, 37(1), 11-15.
Riva, G., Teruzzi, T., & Anolli, L. (2003). The use of the internet in psychological research:
comparison of online and offline questionnaires. CyberPsychology & Behavior, 6( 1), 73-80.
Rotter, J. B. (1954). Social learning and clinical psychology. Englewood Cliffs: Prentice Hall.
Roux, A. V. D. (2013). Conceptual Approaches to the Study of Health Disparities, 41-58. http://doi.org/10.1146/annurev-publhealth-031811-124534.Conceptual
Santiago, C. D., Wadsworth, M. E., & Stump, J. (2011). Socioeconomic status, neighborhood
disadvantage, and poverty-related stress: Prospective effects on psychological syndromes among diverse low-income families. Journal of Economic Psychology, 32(2), 218-230. http://doi.Org/10.1016/j.joep.2009.10.008
Schaefer, J. T., & Magnuson, A. B. (2014). A review of interventions that promote eating by internal cues. Journal of the Academy of Nutrition and Dietetics, 114(5), 734-760. http://doi.Org/10.1016/i.iand.2013.12.024
Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J. (2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. The Journal of educational research, 99(6), 323-338.
Seligman, M. E. (1974). Depression and learned helplessness. InR. J. Friedman & M. M. Katz
(Eds.), The psychology of depression: Contemporary theory and research. Oxford, England:
John Wiley.
Song, J. E., Kim, T., & Ahn, J. A. (2015). A Systematic Review of Psychosocial Interventions for Women with Postpartum Stress. JOGNN - Journal of Obstetric, Gynecologic, and Neonatal Nursing, 44(2), 183-192. http://doi.org/10.llll/1552-6909.12541
Teissedre, F., & Chabrol, H. (2004). Detecting women at risk for postnatal depression using the Edinburgh Postnatal Depression Scale at 2 to 3 days postpartum. Canadian Journal of Psychiatry, 79(1), 51-54.
Thayer, Z. M., & Kuzawa, C. W. (2014). Early origins of health disparities: Material deprivation predicts maternal evening cortisol in pregnancy and offspring cortisol reactivity in the first few weeks of life. American Journal of Human Biology, 26(6), 723-730. http://doi.org/10.1002/ajhb.22532
Thoits, P. A. (2010). Stress and Health: Major Findings and Policy Implications. Journal of Health and Social Behavior, 57(1 Suppl), S41-S53. http ://doi. org/10.1177/0022146510383499


47
Thompson, B., Butcher, A., & Berenson, G. (1987). Children’s beliefs about sources of health: A reliability and validity study. Measurement and Evaluation in Counseling and Development, 20, 80-88.
Tucker, B. (2016). Predictors of a Health Promoting Lifestyle and Behaviors Among Low-Income African American Mothers and White Mothers of Chronically 111 Children.
Ullman, J. B. (2001). Structural equation modeling. InB. G. Tabachnick& L. S. Fidell (Eds.), Using multivariate statistics (4th ed.). Needham Heights, MA: Allyn & Bacon.
Uriu-Adams, J. Y., Obican, S. G., & Keen, C. L. (2013). Vitamin D and maternal and child
health: Overview and implications for dietary requirements. Birth Defects Research Part C -Embryo Today: Reviews, 99( 1), 24-44. http://doi.org/10.1002/bdrc.21031
Wallston, K. A., Wallston, B. S., & DeVellis, R. (1978). Development of the multidimensional health locus of control (MHLC) scales. Health Education & Behavior, 6( 1), 160-170.
Walker, L. O., Cooney, A. T., & Riggs, M. W. (1999). Psychosocial and demographic factors related to health behaviors in the 1st trimester. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 28(6), 606-614.
Walker, S. N., Sechrist, K. R., & Pender, N. J. (1987). The Health-Promoting Lifestyle Profile. Nursing Research, 36(2). doi: 10.1097/00006199-198703 000-00002
Walsh, R. (2011). Lifestyle and Mental. Journal'sdfOvid Full TextAmerican Psychologist, 66(1), 579-592. http://doi.org/10,1037/a0021769
Weitz, T. A., Freund, K. M., & Wright, L. (2001). Identifying and caring for underserved populations: experience of the National Centers of Excellence in Women's Health. Journal of women's health & gender-based medicine, 10(10), 937-952.
Wilkinson, R. G. (1997). Socioeconomic determinants of health. Health inequalities: relative or absolute material standards? BMJ (ClinicalResearch Ed.), 577(7080), 591-595. http://doi.org/10.1136/bmi.314.7080.591
Zimet, Powell, S., Farley, G., Werkman, S., & Berkoff, K. (1990). Psychometric characteristics of the Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 55(3-4), 610-617. http://doi.org/10.1080/00223891.1990.9674095


48
Table 1
APPENDIX
Demographics of Study Sample
Age Category N (%)
15-24 14(8.8)
25-34 88 (55.3)
35-44 54 (34.0)
45-55 2(1.3)
Ethnicity Category N (%)
African American or Black 1(6)
White or Caucasian 133 (83.6)
Latino or Hispanic 17 (10.7)
Asian American or Pacific Islander 3(1.9)
Multiracial 5(3.1)
% Married 89.9%
% Completed HS or above 83%
% Unemployed 16.4%


Table 2
Descriptive statistics for all variables used in mediation analyses
_______________________________Variable_____________________________
Access to Healthcare Resources
Economic Hardship Scale (EHS) composite score
Inability to Make Ends Meet subset total score
Not Enough Money for Necessities subset total score
Adjustments and Cutbacks subset total count
Financial Strain subset total score
Subjective Social Status current (SSS1)
Subjective Social Status future (SSS2)
Multidimensional Scale of Perceived Social Support (MSPSS) Edinburg Postnatal Depression Scale (EPDS)
Learned Helplessness Scale (LHS)
Coping Inventory for Stressful Situations (CISS; total)
Coping Inventory for Stressful Situations (CISS; task oriented)
Health Self-Efficacy Questionnaire (SHEQ; mental health)
Health Self-Efficacy Questionnaire (SHEQ; physical health)
Health Self-Efficacy Questionnaire (SHEQ; total)
Value on Health Scale (VHS)
Multidimensional Health Locus of Control (MHLOC; internal) Multidimensional Health Locus of Control (MHLOC; chance) Multidimensional Health Locus of Control (MHLOC; powerful others) Health Promoting Lifestyle Behaviors (HPLP; total)
49
M SD N
1.09 1.22 159
.18 3.56 125
2.44 1.12 125
2.13 1.00 125
1.52 1.97 125
1.34 .72 125
5.60 1.83 125
6.38 1.79 125
5.91 1.44 129
8.27 5.34 127
8.86 3.04 127
64.16 10.35 127
24.97 5.40 127
57.60 7.63 132
81.74 8.60 132
147.74 16.10 132
15.43 3.10 134
31.28 5.34 126
20.06 6.27 126
18.81 6.26 126
2.63 .50 123


50
Table 3
Correlations of all variables
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1. Age — .12 46** 48** .03 -.003 .35** -.05 -.07 27** .05 -.03 .03 -.14 .04 -.06 .10 .14 .13 .05 -.10 -.02 .25** -.01
2. Employment Status .12 — 48** .08 -.12 -.07 .06 -.12 -.06 .16 .09 -.01 .04 -.10 .00 .02 .10 .04 .06 .04 .11 -.14 .00 -.04
3. Personal Income 46** 48** — .57** .01 -.10 .15 .33** 27** 27** .23* .17 -.12 34** .12 .21* .28** .26** .28** .25** .17 .04 -.01 .30**
4. Family Income 48** .08 .57** — .05 .12 32** 42** .53** .21* .18* .14 -.10 .31** .06 .02 .26** .31* .31** .20* -.01 -.04 -.09 .25**
5. Years of Education .03 -.12 .01 .05 — .05 .004 -.12 -.19* -.07 .07 .004 -.04 .01 .06 .20* .11 .12 .13 .11 .04 .10 .19* .06
6. Relationship Status -.003 -.07 -.10 .12 .05 — -.001 -.01 -.20* .04 .10 .21* -.04 .06 -.04 -.09 .01 -.04 -.02 -.01 -.09 .13 .05 .10
7. Race/Ethnicity .35** .06 -.15 -.32 .004 -.001 — .15 .10 -.08 -.05 -.001 .05 .10 -.04 .11 -.11 -.22* -.18* -.09 -.13 .11 .19* -.23


51
8. Access to Healthcare (higher scores equal less access) -.05 -.12 -.33 -.42* -.12 -.01 .15 — 69** -.06 -.08 32** 34** .25** .02 -.17 48** 44** 49** 23** -.01 .03 -.01 27**
9. EHS composite -.07 -.06 27** .53** 29** .20** .10 g9** — -.16 -.11 .36** 46** 43** .07 -.21* .52** 48** .53** 24** -.11 .12 .02 .36**
10. SSS1 27** .16 27** .21** -.07 .04 -.08 -.06 -.16 — .52** -.02 .11 -.10 -.004 -.08 -.02 -.07 -.05 .12 .04 .07 .07 .02
11. SSS2 .05 .09 23* .18* .07 .10 -.05 -.08 -.11 .52** — .13 .00 -.05 .12 .12 .13 .09 .12 .19* .08 -.02 .04 .20*
12. MSPSS -.03 -.01 .17 .14 .004 .21* -.001 22** .36** -.02 .13 — 37** .30** .15 .28** 48** 40** 47** .21** .05 -.13 .01 42**
13. EPDS .03 .04 -.12 -.10 -.04 -.04 .05 .34* .46* .11 .00 37** — .61 .14 -.23* 69** .62** .68** -.17 -.13 29** .00 49**
14. LHS -.14 -.10 .33** .31** .01 .06 .10 .25** 43** -.10 -.05 30** .61** — .12 29** .54** .53** .56** -.21* -14 39** .10 42**
15. CISS total .04 .00 .12 .06 .06 -.04 -.04 .02 .07 -.004 .12 .15 .14 .12 — .61** .06 .06 .08 -.01 .05 .04 .23* .28**
16. CISS task oriented -.06 .02 .21* .02 .20* -.09 .11 -.17 -.21* -.08 .12 .28** -.23* -.19* .61** — .38** .33** .38** .16 .26** -.18 .13 46**
17. SHEQ mental health .09 .10 .28** .26** .11 .01 -.11 48** .52** -.02 .13 48** 69** .54** .06 .38** — 93** .30** .18 40** -.003 .62**


52
18. SHEQ physical health .14 .04 .26** .31** .12 -.04 -.22* 44** 48** -.07 .09 40** .62** .53** .06 .33** — .95** .36** .19* .33** -.04 .65**
19. SHEQ total .13 .06 .28** .31** .13 -.02 -.18* 49** .53** -.05 .12 47** .68** .56** .08 .38** 93** .95** — 34** .18* 37** -.01 .68**
20. VHS .05 .04 .25** .20* .11 -.01 -.09 23** 24** .12 .19* .21* -.17 -.21* -.01 .16 .30** .36** 34** — 37** .21** .06 49**
21. MHLOC internal -.10 .11 .17 -.01 .04 -.09 -.13 -.01 -.11 .04 .08 .05 -.13 -.14 .05 .26** .18 .19* .18* 37** — -.38 .21 34**
22. MHLOC chance -.02 -.14 .04 -.04 .10 .13 .11 .03 .12 .07 -.02 -.13 29** 39** .04 -.18 40** .33** 37** .21** .38** — .31** .33**
23. MHLOC powerful others .25** .00 -.01 -.09 .19* .05 .19* -.01 .02 .07 .04 .01 .00 .10 .23* .13 -.003 -.04 -.01 .06 .21* .31* — .07
24. HPLP -.01 -.04 .28** .25** .06 .10 -.23* 27** .36** .02 .20* 42** 49** 42** .28** 4g** .62** .65** .68** 49** 39** .33** .07 —


53
Table 4
Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff SE P
X (EHS) .52 .27 .059 .02 .15 .914 -.15 .51 .778 -.19 .26 .475 .01 .02 .755
Ml (MHLOCi) .20 .06 M2 (VHS) — — .57 .36 .120 -.13 .19 .472 .03 .01 .016*
M3 (SHEQ) .09 .05 .118 .01 .00 .001**
M4 (CISSto) — — — — — — — — .02 .01 .005**
Constant 32.79 6.13 <.001 *** 6.007 3.74 .108 133.79 13.15 <001** * 1.31 9.72 .893 -.24 .61 .702
R2= .22 F (14, 92)= 1.83, p = .046* R2= .27 F (15, 91) = 2.27, p = .009** R2= .67 F (16, 90) = 11.42, p< .001*** R2= .30 F (17, 89) = 2.28, p = .007** R2= .65 F (18, 88) = 9.08, p< .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


54
Table 5
Regression of educational achievement on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X (Education) -.06 .75 .934 .38 .39 .345 1.91 1.39 .171 1.06 .71 .138 -.06 .05 .176
Ml (MHLOCi) — — — .20 .06 M2 (VHS) — — — — — — .57 .36 .120 -.13 .19 .472 .03 .01 .016*
M3 (SHEQ) .08 .05 .118 .01 .00 .001**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .005**
Constant 32.79 6.13 * .001***
R2= .22 R2= . 27 R2= .67 R2= .30 R2= . 65
F(14, 92) = 1.83, F (15,91) = 2.27, F (16, 90) = : 11.42, F (17, 89) = 2. 28, F (18, 88) = 9.08,
p = .046 * p = .009** p < .001 *** P = .007** P < .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


55
Table 6
Regression of personal income on maternal lifestyle behaviors with psychological mediators
Antecedent Ml (MHLOCi)
Consequent
M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P
X (Personal Income) .86 .34 .014* .09 .19 .648
Ml (MHLOCi) — — — .20 .06 M2 (VHS) — — — — — —
M3 (SHEQ) — — — — — —
M4 (CISSto) — — — — — —
Constant 32.79 6.13 <001* ** 6.07 3.74 .108
Coeff. SE P Coeff. SE P Coeff. SE P
-.65 .65 .320 .54 .33 .109 .04 .02 .053
.35 .21 .087 .21 .11 .047 .01 .01 .141
.57 .36 .120 -.13 .19 .472 .03 .01 .016*
_ _ _ .08 .05 .118 .01 .00 .001**
.02 .01 .005**
133.79 13.11 < 1.31 9.71 .893 -.24 .61 .702
.001***
R2= .22
F (14, 92)= 1.83, p = .046*
R2= .27
F (15, 91) = 2.27, p = .009**
R2= .67
F (16, 90)= 11.42,
p< .001***
R2= .3
F (17, 89) = 2.28, p= ,007*(
R2= .65
F (18, 88) = 9.09,
p < .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level


56
Table 7
Regression of family income on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X (Family -.27 .42 .523 .19 .22 .384 -.18 .77 .819 -.19 .39 .625 .02 .03 .385
Income) Ml (MHLOCi) .20 .05 M2 (VHS) — — — — — — .58 .36 .119 -.13 .19 All .03 .01 .016*
M3 (SHEQ) .09 .05 .118 .01 .00 .001**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .005**
Constant 32.79 6.13 <001* 6.07 3.74 .108 133.79 11.42 **
R2= .22
F (14, 92)= 1.83, p = .046*
R2= .27
F (15, 91) = 2.27, p = .009**
R2= .67
F (16, 90)= 11.42,
p < 001***
R2= .30
F (17, 89) = 2.28, p = .007**
R2= .65
F (18, 88) = 9.08,
p< .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


57
Table 8
Regression of social support on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS)
M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P
X (Social Support) .38 .54 .476 .09 .29 .742
Ml (MHLOCi) — — — .20 .05 M2 (VHS) — — — — — —
M3 (SHEQ) — — — — — —
M4 (CISSto) — — — — — —
Constant 32.79 6.13 <001* ** 6.07 3.74 .108
Coeff. SE P Coeff. SE P Coeff. SE P
1.10 .99 .269 .53 .50 .292 .02 .03 .530
.36 .21 .086 .21 .11 .047* .01 .01 .141
.57 .36 .120 -.13 .19 .472 .03 .01 .016*
.08 .05 .118 .01 .00 .001**
.02 .01 .005**
133.79 13.11 R2= .22
F (14, 92)= 1.83, p = .046*
R2= .27
F (15, 91) = 2.27, p = .009**
R2= .67
F (16, 90)= 11.42,
p < .001***
R2= .30
F (17, 89) = 2.28, p = .007**
R2= .65
F (18, 88) = 9.08,
p < .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


58
Table 9
Regression of Subjective Social Status (SSS1; current) on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X(SSS1) -.53 .42 .203 .08 .22 .709 .35 .77 .654 -.42 .39 .279 -.01 .03 .577
Ml (MHLOCi) — — — .20 .06 M2 (VHS) — — — — — — .57 .36 .120 -.13 .18 .472 .03 .01 .016*
M3 (SHEQ) .08 .05 .118 .01 .00 .001**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .005**
Constant 32.79 6.13 <001 6.07 3.74 .108 133.79 11.42 < 1.31 9.72 .893 -.24 .61 .702
*** 001***
R2= .22
F (14, 92)= 1.83, p = .046*
R2= .27
F (15, 91) = 2.27, p = .009**
R2= .67
F (16, 90)= 11.42,
p < 001***
R2= .30
F (17, 89) = 2.28, p = .007*
R2= .65
F (18, 88) = 9.08,
p< .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


59
Table 10
Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X (SSS2) .45 .40 .266 .18 .21 .402 .58 .74 .434 .35 .38 .361 .01 .02 .823
Ml (MHLOCi) — — — .20 .06 M2 (VHS) — — — — — — .57 .36 .119 -.13 .19 All .03 .01 .016*
M3 (SHEQ) .08 .05 .118 .01 .00 .001**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .005**
Constant 32.79 6.13 <001 6.07 3.74 .108 133.79 13.11
R2= .22
F (14, 92)= 1.83, p = .046
R2= .27
F (15, 91) = 2.27, p = .009**
R2= .67
F (16, 90)= 11.42,
p < 001***
R2= .30
F (17, 89) = 2.28, p = .007**
R2= .65
F (18, 88) = 9.08,
p< .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


60
Table 11
Regression of access to health resources on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE P Coeff. SE P
X (Access to Healthcare) 1.18 .66 .078 -.13 .36 .713 -1.76 1.24 .159 .37 .63 .559 .04 .04 .338
Ml (MHLOCi) — — — .20 .06 <001*** .36 .21 .087 .21 .11 .047 * .01 .01 .141
M2 (VHS) — .57 .36 .119 -.13 .19 .472 .03 .01 .016*
M3 (SHEQ) .08 .05 .118 .01 .00 .001**
M4 (CISSto) — — — — .02 .01 .005**
Constant 32.79 6.13 <001* 6.07 3.74 .108 133.79 11.42 < 1.31 9.72 .893 -.24 .61 .702
** .001***
R2= .22 R2= .27 R2= .67 R2 = .30 R2= .65
F (14, 92)= 1.83, F (15, 91) = 2.27, F (16, 90)= 11.42, F (17, 89) = 2.28, F (18, 88) = = 9.08,
p = .046* p = .009** p < .001*** P = .007** P < .001 ***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level


61
Table 12
Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with maternal
depression and learned helplessness mediators________________________________________
Consequent
Antecedent Ml (EPDS) M2 (LHS) Y (HPLP)
Coeff. SE p Coeff. SE P Coeff. SE P
X (EHS) .47 .19 .012* .05 .12 .694 .01 .02 .755
Ml (EPDS) — — — .19 .07 .006** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.31 5.67 < 18.25 3.99 <.001* -.24 .61 .702
.001*** **
R2= .63 R2= .55 R2= .65
F (16, 90) = = 9.67, F (17, 89) = = 6.40, F (18, 88) = = 9.08,
p < .001 *** R. < .001 *** R. < .001 ***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Table 13
Regression of educational achievement on maternal lifestyle behaviors with maternal depression
and learned helplessness mediators_________________________________________________________
Consequent
Antecedent
Ml (EPDS)
M2 (LHS)
Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P
X (Yrs. 1.04 .52 .047* -.13 .33 .695 -.06 .05 .178
Education) Ml (EPDS) .19 .04 .006** -.01 .01 336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.67 < 18.25 3.99 < .001*** -.23 .61 .702
R2= . F (16, 90) .001*** 63 = 9.67, R2= . F (17, 89) 55 = 6.81, R2= . F (18, 88) 65 = 9.08,


62
p < .001***
p < .001***
p< .001***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Table 14
Regression of personal income on maternal lifestyle behaviors with maternal depression and
learned helplessness mediators____________________________________________________________
Consequent
Antecedent Ml (EPDS)
M2 (LHS) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE P
X (Personal -.20 .24 .412 -.22 .15 .152 .04 .02 .053
Income) Ml (EPDS) .19 .07 .006** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.84 < .001*** 18.25 3.99 <001* -.24 .61 .702
**
R2= .63 R2= .55 R2= .65
F (16, 90) = 9.67,
p< ,001***
F (17, 89) = 6.40, F (18, 88) = 9.08,
p < ,001***_____________________p< .001***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Table 15
Regression of family income on maternal lifestyle behaviors with maternal depression and
learned helplessness mediators_____________________________________________________
Consequent
Antecedent Ml (EPDS) M2 (LHS) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P
X (Family .27 .28 .328 -.39 .17 .026 .02 .03 .385
Income)


63
Ml (EPDS) — — — .19 .07 .006** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.67 <.001*** 18.25 4.57 <001* ** -.24 .61 .702
R2= .63 R2= .55 R2= . 65
F (16, 90) = 9.67, F (17, 89) = = 6.40, F (18, 88) = 9.08,
p< .001*** R < .001 *** R < .001***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Table 16
Regression of social support on maternal lifestyle behaviors with maternal depression and
learned helplessness mediators____________________________________________________________
Consequent
Antecedent Ml (EPDS) M2 (LHS) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE P
X (Social -.61 .36 .101 -.35 .23 .129 .02 .03 .631
Support) Ml (EPDS) .19 .06 .006** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.67 < .001*** 18.25 3.99 <001* 5k 5k -.24 .61 .702
R2= . 63 R2= .55 R2= . 65
F (16, 90) = 9.67, F (17, 89) = = 6.40, F (18, 88) = 9.08,
p< .001*** R < .001 *** p< .001***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


64
Table 17
Regression of Subjective Social Status (SSS1; current) on maternal lifestyle behaviors with
maternal depression and learned helplessness mediators____________________________
Consequent
Antecedent Ml (EPDS) M2 (LHS) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE P
X(SSS1) .68 .28 .018* .00 .18 .998 -.01 .03 .577
Ml (EPDS) — — — .19 .07 .064** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.67 <. 001*** 18.25 3.99 <001* ** -.24 .61 .702
R2= .63 R2= .55 R2= .65
F (16, 90) = 9.67, F (17, 89) = 6. 40, F (18, 88) = 9. 08,
p< .001*** p < .001** p< .001***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Table 18
Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with
maternal depression and learned helplessness mediators____________________________
Consequent
Antecedent Ml (EPDS) M2 (LHS) Y (HPLP)
Coeff. SE P Coeff. SE P Coeff. SE P
X (SSS2) -.09 .28 .741 .05 .17 .778 .01 .02 .823
Ml (EPDS) — — — .19 .06 .006** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.67 <. 001*** 18.25 3.99 <001* ** -.24 .61 .702
R2= .63 R2= .55 R2= .65
F (16, 90) = 9.67, F (17, 89) = 6 .40, F (18, 88) = 9. 08,


65
p < .001***
p< .001***
p< .001***
Note. aEPDS = Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Table 19
Regression of access to health resources on maternal lifestyle behaviors with maternal
depression and learned helplessness mediators_____________________________________________
Consequent
Antecedent Ml (EPDS) M2 (LHS) Y (HPLP)
Coeff SE p Coeff. SE P Coeff. SE P
X (Access to -.29 .46 .527 -.45 .29 .121 .04 .04 .338
Healthcare) Ml (EPDS) .19 .06 .006** -.01 .01 .336
M2 (LHS) — — — — — — -.01 .02 .626
Constant 29.32 5.67 <.001*** 18.25 3.99 <001** * -.24 .61 .702
R2= .63 R2= . 55 R2= .64
F (16, 90) = 9.67, p< .001** F (17, 89) = 6.40, p< .001*** F (18, 88) = 9.08, p< .001***
= Edinburgh Postnatal Depression Scale, bLHS = Learned helplessness scale (shortened), CHPLP = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


66
Table 20
Regression of inability to make ends meet subscale ofEHS on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X (Inability to -.15 1.30 .909 -.01 .69 .991 2.00 2.39 .404 .34 1.21 .111 .09 .08 .263
Make Ends
Meet)
Ml (MHLOCi) — — — .20 .05 M2 (VHS) — — — — — — .57 .36 .120 -.13 .19 All .03 .01 .015*
M3 (SHEQ) .09 .05 .128 .01 .00 .002**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .005**
Constant 33.18 7.04 <001* 6.09 4.18 .149 128.50 14.57 < .58 10.10 .954 -.42 .64 .512
** .001***
R2= .22 R2= . 27 R2= .67 R: * = .30 R2= .66
F (15, 91)= 1 .69, F (16, 90) = 2.11, F (11 ii 00 : 10.76, â–ºn oo 88) = 2. 13, F (19, 87) = = 8.69,
p = .067 p = .014* P < .001 *** p = = .011* P < .001 ***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


67
Table 21
Regression of adjustments and cutbacks subset of EHS on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X (Adjustments -.20 .60 .736 .04 .32 .907 .20 1.11 .858 -.84 .55 .133 .04 .04 221
and Cutbacks)
Ml (MHLOCi) — — — .21 .06 M2 (VHS) — — — — — — .57 .36 .122 -.13 .18 All .03 .01 .016*
M3 (SHEQ) .09 .05 .109 .01 .00 .002**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .003**
Constant 33.02 6.20 <001* 6.02 3.78 .115 133.53 13.26 < 2.22 9.66 .819 -.28 .61 .645
** .001***
R2= .22 R2= . 27 R2= .67 R2 = .32 R2= .66
F (15, 91)= : 1.70, F (16, 90) = 2.11, F (17, 89) = : 10.64, F (18, 88) = 2.31, F (19, 87) = = 8.73,
p = .065 p = .014* P < .001 *** P = .005** P < .001 ***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level


68
Table 22
Regression of not enough money for necessities on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE P Coeff. SE P Coeff. SE P
X (Not Enough -.68 1.35 .618 -.44 .72 .539 -1.43 2.49 .569 1.69 1.25 .19 -.22 .08 .006**
for Money Necessities) Ml (MHLOCi) .20 .05 M2 (VHS) — — — — — — .56 .36 .131 -.12 .19 .517 .03 .01 .017*
M3 (SHEQ) .09 .05 .099 .01 .00 .002**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .001**
Constant 34.35 6.90 <001* 7.15 4.14 .088 137.34 14.55 < -3.49 10.30 .735 .38 .63 .544
** 001***
R2= .22
F (15, 91)= 1.71, p= .063
R2= .27
F (16, 90) = 2.14, p= .013*
R2= .67
F (17, 89)= 10.69,
p < 001***
R2= .32
F (18, 88) = 2.27,
p = .006**
R2= .68
F (18, 87) = 9.72,
p< .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level


69
Table 23
Regression of financial strain subset on maternal lifestyle behaviors with psychological mediators
Consequent
Antecedent Ml (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP)
Coeff SE P Coeff. SE P Coeff. SE p Coeff. SE P Coeff. SE P
X (Financial -.88 1.29 .498 .23 .69 .737 -.67 2.37 .778 .09 1.20 .938 .02 .08 .847
Strain)
Ml (MHLOCi) — — — .20 .06 M2 (VHS) — — — — — — .57 .36 .119 -.11 .19 .564 .03 .01 .017*
M3 (SHEQ) .09 .05 .120 .01 .00 .001**
M4 (CISSto) — — — — — — — — — — — — .02 .01 .005**
Constant 31.70 6.36 <001* 5.83 3.83 .131 134.48 13.39 < 1.20 9.88 .904 -.25 .62 .686
** .001***
R2= .22 R2= . 27 R2= .67 R2 = .30 R2= .65
F (15, 91) = 1.73 F (16, 90) = 2.12, F (11 ii 57 00 : 10.65, F (18, 88) = 2. 13, F (189, 87) = 8.51,
p = .059 p = .014* P < .001 *** P = .011* P < .001***
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-
efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eHPLP = Health Promoting Lifestyle
Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level


Model 1
Model 2
70
Model 3
Figure 1. Diagram of Conceptual framework of three models.
Note. aMHLOCi = Multidimensional health locus of control, internal subscale, bVHS = Value on health scale, CSHEQ = School health self-efficacy questionnaire, dCISSto = Coping inventory for stressful situations, task oriented subscale, eEPDS = Edinburgh Postnatal Depression scale, rLHS = Learned Helplessness Scale, gHPLP = Health Promoting Lifestyle Profil


71
.21*
c = -.01
Figure 2. Diagram of psychological mediators on relationship between economic hardship and maternal lifestyle behaviors.
Note. aX = Economic Hardship Scale, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.
.21
c = -.01
Figure 3. Diagram of psychological mediators on relationship between educational achievement and maternal lifestyle behaviors.
Note. aX = Years of education, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


72
.21
c = .07
Figure 4. Diagram of psychological mediators on relationship between annual personal income and maternal lifestyle behaviors.
Note. aX = Annual personal income, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.
.21*
c = ,01
Figure 5. Diagram of psychological mediators on relationship between annual family income and maternal lifestyle behaviors.
Note. aX = Annual family income, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile


73
*Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.
.21*
c - .06
Figure 6. Diagram of psychological mediators on relationship between social support and maternal lifestyle behaviors.
Note. aX = Multidimensional Scale of Perceived Social Support, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
*Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.
.21*
c = -.03
Figure 7. Diagram of psychological mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors.
Note. aX = Subjective Social Status, current, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire,


74
eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
.21*
c = .03
Figure 8. Diagram of psychological mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors.
Note. aX = Subjective Social Status, future, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 = Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
.21*
c= .05
Figure 9. Diagram of psychological mediators on relationship between access to health resources and maternal lifestyle behaviors.
Note. aX = Access to health resources, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dM3 = School health self-efficacy questionnaire, eM4 =


75
Coping inventory for stressful situations, task oriented subscale, fY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Figure 10. Diagram of maternal depression and learned helplessness mediators on relationship between economic hardship and maternal lifestyle behaviors.
Note. aX = Economic Hardship Scale, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
a
c = -.07
Figure 11. Diagram of maternal depression and learned helplessness mediators on relationship between educational achievement and maternal lifestyle behaviors.
Note. aX = Years of education, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


76

Figure 12. Diagram of maternal depression and learned helplessness mediators on relationship between annual personal income and maternal lifestyle behaviors.
Note. aX = Annual personal income, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Figure 13. Diagram of maternal depression and learned helplessness mediators on relationship between annual family income and maternal lifestyle behaviors.
Note. aX = Annual family income, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


77
a
Figure 14. Diagram of maternal depression and learned helplessness mediators on relationship between social support and maternal lifestyle behaviors.
Note. aX = Multidimensional Scale of Perceived Social Support, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Figure 15. Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors.
Note. aX = Subjective Social Status, current, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


78
Figure 16. Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors.
Note. aX = Subjective Social Status, future, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.

IQ**
Figure 17. Diagram of maternal depression and learned helplessness mediators on relationship between access to health resources and maternal lifestyle behaviors.
Note. aX = Access to health resources, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


79
.21*
Figure 18. Diagram of psychological mediators on relationship between inability to make ends meet subset of EHS and maternal lifestyle behaviors.
Note. aX = Inability to make ends meet average, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.
Figure 19. Diagram of psychological mediators on relationship between adjustments and cutbacks subset of EHS and maternal lifestyle behaviors.
Note. aX = Adjustments and cutbacks count, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.


80
Figure 20. Diagram of psychological mediators on relationship between not enough money for necessities of EHS and maternal lifestyle behaviors.
Note. aX = Not enough money for necessities average, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, * Correlation is significant at .01 level, ** Correlation is significant at <.001 level.
.21*
c = .04
Figure 21. Diagram of psychological mediators on relationship between financial strain subset of EHS and maternal lifestyle behaviors.
Note. aX = Financial strain average, bMl = Multidimensional health locus of control, internal subscale, CM2 = Value on health scale, dY = Health Promoting Lifestyle Profile Correlation is significant at .05 level, ^correlation is significant at .01 level, **Correlation is significant at <.001 level.


81
Figure 22. Diagram of SEM model makeup of latent construct “socioeconomic resources” and prediction of maternal lifestyle behaviors.
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.
Figure 23. Diagram of SEM of latent construct “socioeconomic resources” prediction of maternal lifestyle behaviors through mediator maternal depression.
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


82
Figure 24. Diagram of SEM of latent construct “socioeconomic resources” prediction of maternal lifestyle behaviors through mediators of binary variables for years of education and race/ethnicity.
Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.


Full Text

PAGE 1

i THE EFFECTS OF DEMOGRAPHIC AND SOCIOECONOMIC FACTORS ON LIFESTYLE BEHAVIORS IN MOTHERS by KATHRYN SCHEYER B.A. , California State University, Long Beach, 2015 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Clinical Health Psychology Program 2018

PAGE 2

ii This thesis for the Master of Arts degree by Kathryn Scheyer has been approved for the Clinical Health Psychology Program by Kevin Everhart, Chair, Advisor Peter Kaplan, Co Advisor Jonathan Shaffer Zaneta Thayer Date: July 28 , 2018

PAGE 3

iii Scheyer, Kathryn B.A., (MA, Clinical Health Psychology Program) The Effects of Demographic and Socioeconomic Factors on Lifestyle Behaviors in Mothers Thesis directed by Associate Professor, Kevin Everhart ABSTRACT Health disparities are health outcomes t hat vary based on factors such as race, ethnicity or sociodemographic variables. These differences in health are often associated with lower income and lower educational status which can lead to fewer health care resources and p oorer health. The current s tudy will examine one mechanism through which risk factors lead to health disparities namely, through the lack of engagement in lifestyle promoting behaviors . The current study investigate d : (1) the effect s of specific socioeconomic and sociodemographic ri sk factors on the health promoting lifestyle behaviors of mothers, and (2) the psychological factors that influence th e s e relationship s . T here is research to support a Health Self Empowerment theoretical model that predicts that some mothers who experience socioeco nomic and sociodemographic risk factors will engage in health promoting lifestyle behaviors because of individual psycholo gical variables. The current study found no significant direct mediation ef fects of psychological factors on the relationship between specific demographic and socioeconomic factors on maternal lifestyle behaviors. However, significant indirect pathways were found between some variables investigated, including the relationships between economic hardship, educa tional achievement, personal income, and perception of social status and the pathways of internal health locus of control, health motivation, health self efficacy and task oriented coping, leading to higher maternal engagement in health promoting lifestyle behaviors . The form and content of this abstract are approved. I recommend its publication. Approved: Kevin Everh art

PAGE 4

iv TABLE OF CONTENTS I. INTRODUCTION . Lifestyle Behavior and Healt h .. . .. Psychological .. Research II. METHOD .. . ... .. Demographic questionnair e Active Health self

PAGE 5

v .. .. III. RESULTS .. ... Sample ... ... .32 Post Hoc IV. DISCUSSION ... .. .42 Study REFERENCES

PAGE 6

vi LIST OF TABLES TABLE 1. Demographics of Study Sample 2. Descriptive statistics for all variables used in mediation analyses 3. Correlations of all variables 4. Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with psychological mediators 5. Regression of educational achievement on maternal lifestyle behaviors with psychological mediators 6. Regression of personal income on maternal lifestyle behaviors with psychological mediators 7. Regression of family income on maternal lifestyle behaviors with psychological mediators 8. Regression of social support on maternal lifestyle behaviors with psycholog ical mediators 9. Regression of Subjective Social Status (SSS1; current) on maternal lifestyle behaviors with psychological mediators 10. Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with psychological mediators 11. Regression of access to health resources on maternal lifestyle behaviors with psychological mediators 12. Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators

PAGE 7

vii 13. Regression of educational ac hievement on maternal lifestyle behaviors with maternal depression and learned helplessness mediators 14. Regression of personal income on maternal lifestyle behaviors with maternal depression and learned helplessness mediators 15. Regression of family income on m aternal lifestyle behaviors with maternal depression and learned helplessness mediators 16. Regression of social support on maternal lifestyle behaviors with maternal depression and learned helplessness mediators 17. Regression of Subjective Social Status (SSS1; c urrent) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators 18. Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators 19. Regression of access to health resources on maternal lifestyle behaviors with maternal depression and learned helplessness mediators 20. Regression of inability to make ends meet subscale of EHS on maternal lifestyle behaviors with psychological mediators 21. Regression of adjustments and cutbacks subset of EHS on maternal lifestyle behaviors with psychological mediators 22. Regression of not enough money for necessities on maternal lifestyle behaviors with psychological mediators 23. Regression of financial strain subset on maternal lifestyle behaviors with psychological mediators

PAGE 8

viii LIST OF FIGURES FIGURE 1. Diagram of Conceptual framework of three models. 2. Diagram of psychological mediators on relationship between economic hardship and maternal lifestyle behaviors. 3. Diagram of psy chological mediators on relationship between educational achievement and maternal lifestyle behaviors. 4. Diagram of psychological mediators on relationship between annual personal income and maternal lifestyle behaviors. 5. Diagram of psychological mediators on relationship between annual family income and maternal lifestyle behaviors. 6. Diagram of psychological mediators on relationship between social support and maternal lifestyle behaviors. 7. Diagram of psychological mediators on relationship between Subjective S ocial Status, current, and maternal lifestyle behaviors. 8. Diagram of psychological mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors. 9. Diagram of psychological mediators on relationship between access to hea lth resources and maternal lifestyle behaviors. 10. Diagram of maternal depression and learned helplessness mediators on relationship between economic hardship and maternal lifestyle behaviors. 11. Diagram of maternal depression and learned helplessness mediators on relationship between educational achievement and maternal lifestyle behaviors.

PAGE 9

ix 12. Diagram of maternal depression and learned helplessness mediators on relationship between annual personal income and maternal lifestyle behaviors. 13. Diagram of maternal depres sion and learned helplessness mediators on relationship between annual family income and maternal lifestyle behaviors. 14. Diagram of maternal depression and learned helplessness mediators on relationship between social support and maternal lifestyle behaviors . 15. Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors. 16. Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors. 17. Diagram of maternal depression and learned helplessness mediators on relationship between access to health resources and maternal lifestyle behaviors. 18. Diagram of psychological mediators on relationship between inability to make ends meet subset of EHS and maternal lifestyle behaviors. 19. Diagram of psychological mediators on relationship between adjustments and cutbacks subset of EHS and maternal lifestyle behaviors. 20. Diagram of psychological m ediators on relationship between not enough money for necessities of EHS and maternal lifestyle behaviors. 21. Diagram of psychological mediators on relationship between financial strain subset of EHS and maternal lifestyle behaviors. 22. Diagram of SEM model make prediction of maternal lifestyle behaviors. 23. lifestyle behaviors through mediator maternal depression.

PAGE 10

x 24. Diagram of SEM of lifestyle behaviors through mediators of binary variables for years of education and race/ethnicity.

PAGE 11

1 CHAPTER 1 INTRODUCTION The Effect of Demographic and Socioeconomic Risk Factor s on Life Style Behaviors in Mothers Health disparities are health outcomes that vary based on rac ial identity , ethnic differences, or sociodemographic variables. Understanding the underlying risk factors for health disparities is a major public health con cern. One hypothesis is that health disparities result from differences in stress exposures and environment experience s of disadvantaged populations , as race, socioeconomic status and health have been historically intertwined in the United States (Fiscella & Williams , 2004). In addition, l ower income and education often lead to less access to health care resources and the development of poor health . T he current study examin ed an additional mechanism through which health disparities may occur in contemporary society : differential engagement in lifestyle promoting behaviors . I specifically evaluate d the relationships among socio demographic and socio economic factors and engagement in health promoting lifestyle behaviors by mothers , and (2) the psychological factors that influence th ese relationship s . Lifestyle Behavior and Health H ealth promoting behaviors, such as exercise, practicing good nutrition and utilizing stress management techniques, have been linked to better mental and physical health outcomes across populations. Physical activity has been found to be preventative for di sability in aging populations (Buford, Anton, Clark, Higgins, & Cooke 2014) , beneficial for improving overall health (Mitchell & Barlow, 2011) and preventing and/or treating numerous m edical conditions (Walsh , 2011). A five y ear review on exercise and qual ity of life (QOL) found a general consensus in the literature that higher levels of exercise result in higher QOL in both health y

PAGE 12

2 individuals and those with specific ailments, such as those found in aging populations (Mitchell & Barlow, 2011). Nutrition ma y also be considered a modifiable risk factor for some psychopathology (Low Dog, 2010) , with implications for both physical and mental health . Overall, improvements in eating habits and lifestyle, including dieting and physical activity, are significantly positively associated with long term improvements in mental health, such as depression, anxiety, self esteem and negative affect . Higher levels of physical activity and healthy eating habits have resulted in improved QOL and improved psychological health (including a reduction in depression and anxiety and improvements in self esteem) for adult populations, children and adolescents (Biddle & Asare, 2011; Schaefer & Magnuson, 2014 ). Healthy eating and engaging in physical activity enhan ce cognitive functioning, academic achievement among adults, children and adolescents and reduce age related memory loss and the risk of dementia in elderly populations (Biddle & Asare, 2011; Walsh, 2011). Additionally, stress management techniques, inclu ding mindfulness based stress reduction decrease s stress among populations with multiple physical ailments and psychological difficulties, including medical symptoms, pain, physical impairment, depression, and anxiety (Grossman , Niemann, Schmidt & Walach 2 004). A review article on lifestyle and mental health describes therapeutic lifestyle changes (TLC), including exercise, diet and nutrition, recreation, relaxation and stress management and religious or spiritual involvement, as a central focus in overall mental and physical health (Walsh, 2011). Dem ographic and Socioeconomic Factors and Maternal Health Given that previous research provides a link between engaging in health promoting lifestyle behaviors and better mental and physical health (Prather, Spitznagle, & Hunt , 2012), it is important to understand the relationship s among risk factors and the health of mothers and

PAGE 13

3 infants . Previous research on the health of mother and infant population s has show n that early life exposure to so cial disadvantages or unfavorable health conditions have potential long lasting effects. Maternal chronic stress, which can be shaped by socioeconomic and sociodemographic factors such as low socioeconomic status ( SES ) , low social and partner support, low educational achievement, and low access to resources, significantly jeopardize the health of both mother and baby ( Ghosh, Willhelm, Dunkel Schetter, Lombardi, & Ritz, 2010; Lancaster, et al., 2010; Lueke n et al., 2008 ; Matijasevich et al., 2010; Premji, 2014; Thayer & Kuzawa 2014). Matijasevich et al. (2010) assessed health outcomes of socioeconomic factors that contribute to health disparities through multiple birth cohort studies in Brazil and the United Kingdom. They found that mothers from poorer and less educated backgrounds had more adverse health behaviors and outcomes, such as smoking during pregnancy, preterm birth and shortened duration of breastfeeding practices (breastfed infant for less than 3 months) . S ocio demographic factors such as income, educational attainment and partner status have also been found to be predictive factors of depression symptoms during and after pregnancy . Social factors, specifically social support and marital relationships , also influence the perinatal mental health of women (Premji, 2014) and a decrease in social support is negatively associated with maternal health problems. Through the large scale Fragile Families and Child Wellbeing Study (N = 12,140), Har k nett & Har t nett (20 11) found that l ess social support and lack of personal safety nets through social networks were predicted by low er income and further, mothers with poor physical and mental health (higher rates of depression) were less likely to have financial, housing or child care support available to them . In addition to experiencing prenatal stress and health problems, m others from relatively low income backgrounds have historically received poor er maternal care. Abel (1996) studied

PAGE 14

4 the relationship of socio economic f actors such as ethnicity (African American versus White/ Caucasian) , education (less than a high school education versus high school education) and marital status (married versus unmarried) with the receipt of adequate prenatal care among 11,936 birth case s and found that these risk factors significantly predicted inadequate prenatal care. L ower income, fewer years of education and residential property status predicted higher scores o n the Edinburg Postnatal Depression Scale (EPDS) and that partner status ( i.e., not having a partner) and family support w ere most predictive of depressive symptomology in mothers (Hein et al., 2014; Lancaster et al., 2010) . Dem ographic and Socioeconomic Factors and Child Health M aternal and child health are interrelated . P oor maternal health and elevated maternal stress , due to low income and other sociodemographic risk factors, is predictive of both adverse birth outcomes (Doming uez, 2011 ; Gennaro & Hennessy , 2003 ; Hobel, Dunkel Schetter, Roesch, Castro & Arora, 1999 ) and poorer child health outcomes (Minkovitz, & Grason, 2002; Hardie & Landale , 2013). Studies have shown a link between prenatal maternal stress, which seems to be pronounced in low SES and ethnic minority families (Luecken et al., 2010 ) and adv erse infant health outcomes, including preterm birth ( Dole & Mollborn , 2003 ), low birth weight (Dennis et al., 2013 ; Parker, Schoendorf, & Kiely, 1994), and infant mortality (James, 1993 ). Consequently , inadequate prenatal care predicts low birth weight an d infant mortality (Gortmarker, 1979). There is also research that supports the relationship between poverty (income and maternal educational achievement) , child stress physiology , and child health and development outcomes after infancy . Socioeconomic status levels and environmental threat and hostility , which contribute to maternal stress, may have long lasting effects on the developing hypothalamic pituitary adrenal

PAGE 15

5 ( HPA ) regulating system . Children who grow up in hostile or threatening environments m ay develop a highly sensitive H PA system in response to environmental stress and the absence of socioeconomic resources (e.g., social support) ( Everhart & Emde, 2006 ) . Ph ysical and psychosocial stressors, such as housing quality and violence were found to be associated with heightened cardiovascular and neuroendocrine levels and cumulative stress in early childhood (Evans , 2003 ; Everhart & Emde, 2006 ). Additionally , cultural and environmental context, such as low SES and maternal stress significantly predict higher infant cortisol reactivity after birth ( Thayer et al. 2014) and higher salivary cortisol levels in infants and at 4.5 years of age (Clearfield et al., 2014; Essex, Klein, Cho & Kalin, 2002). Furthermore, s ocioeconomic influence on HPA regula tory systems may be responsible for dysregulation in early life and child psychopathology (Everhart & Emde, 2006 ). Elevated salivary cortisol and reactivity has potential long term effects o n child biology and health through a variety of psychological and physical health problems such as anxiety disorders, depression, somatic complaints, aggression, attention problems, cardiovascular disease, respiratory disease and some types of cancer ( McEwan, 2007; Miller et al., 2009; Santiago, Wadsworth & Stump, 2011) . Lifestyle Behaviors and Maternal/Child Health Health promoting behaviors are particularly important in the area of maternal and infant health, as previous research has established a relationship between healthy lifestyle behaviors and the mental and physical health of both mother and infant . Prather et al., (2012) for instance, reviewed previous studies showing that exercise and appropriate nutrition are important contributors to maternal and infant physical and psychological health. The ir review identifies maternal benefits of exercise , including improved cardiovascular function, lower risk for gestational diabetes, improved strength and lean muscle mass, improved sense of well being , and

PAGE 16

6 enhanced sleep . It further found that exercise h as a n ti depressant effects, improves self esteem and reduces symptoms of anxiety and depression during pregnancy. M aternal exercise during pregnancy also affect s infant health, resulting in higher birth weights and potentially improved neurodevelopment (improv ed orientation, ability to self sooth in neonates and higher general intelligence and oral language scores at age five ; Prather et al., 2012). Maternal nutrition , including appropriate diet with sufficient vitamin D and weight management, is also an import ant factor that plays a role in shaping the health of mothers and infants ( Prather et al., 2012; Uriu Adams, Obican & Keen, 2013 ). R elaxation techniques and other stress management interventions positively affect maternal health. Improvements in health include i mproved emotional state during pregnancy, improved pregnancy outcomes (e.g., fewer admissions to the hospital, fewer postpartum complications), improved fetal and neonatal outcomes , including a reduction in fetal heart rate and fetal motor activit y, a reduction in maternal physiological and endocrine measures (Fink , Urech, Cavelti, & Adler , 2012) and reduced postpartum stress in general (Song, Kim & Ahn, 2015). S tress researchers have found that individuals from low income or low educational backgr ounds have the highest rates of morbidity, disability, mortality, psychological distress and mental disorders compared to those from more advantageous socioeconomic backgrounds (Thoit s , 2010 ). Socioeconomic and sociodemographic r isk factors affect a variet y of lifestyle behaviors that prom ote physical and mental health and s pecific resources can help individuals deal with demands and cumulative psychosocial stressors. Mothers from low income backgrounds may not possess the coping resources, knowledge, money, or access to healthcare that mothers from higher SES backgrounds have through which to secure good health (Thoit et

PAGE 17

7 al., 2010; Roux et al., 2013). Indeed, p revious research has shown that education and income are positively associated wit h health promotion practices in general (practicing good nutrition, stress managemen t; Duffy , 1997 ; Walker , Sechrist , & Pender 198 7 ) and negatively associated with the risk of weight gain ( Kahn, Williamson, & Stevens 1991 ). Additionally, Duffy et al. (1997 ) found that specific demographic factors, such as educational level, were associated with improved frequency of engaging in health promoting activities including exercise, nutrition, stress management, interpersonal support, health responsibility and self actualization in mothers . Other literature also suggests that mothers from low income and minority backgrounds have less access to economic resources and knowledge that promotes healthy lifestyle behaviors such as exercise and stress management. In underserved populations (such as populations of low SES, racial or ethnic minorities or non English speaking; Weitz, Freund & Wright, 2001) access to maternal health resources and factors such as time and money serve as barriers for women to engage in these specific hea l th promoting behaviors (Duffy et al., 1997 ; Gazmararian , Adams, & Pamuk 1995 ). Gazmararian et al. (1995) found specific sociodemographic risk factors such as education level, poverty and Medicaid status, predict some maternal lifestyle b ehaviors such as smoking. Specific health behaviors detrimental to health outcomes, known as health risk behaviors (e.g., smoking) , are also more prevalent in low income populations, specifically among those with low SES and educational status that are par ticularly harmful to maternal and in fant health outcomes (Gazmararian et al., 1995; Lantz et al., 2001 ). Psychological Mediators Given findings that show a consistent relationship between demographic and socioeconomic risk factors and maternal and infant health, and the positive relationship between engagement in health promoting lifestyle behaviors and overall health, the current study aimed

PAGE 18

8 to understand the specific relationship between these socioeconomic and sociod emographic factors and lifestyle behaviors in mothers and what factors may mediate this relationship . I hypothesize that m individual atti tudes about self and health will mediate the association of socioeconomic and sociodemographic factors with the engagement in health promoting lifestyle behaviors. Specific psychological variables may mediate the relationship between risk factors and lifestyle behaviors and explain why some women engage in health promoting lifestyle behaviors. For instance, research has shown that psychological variables serve as predictive factors of health promoting lifestyle behaviors. A review by Matthews, Gallo and Taylor in 2010 proposes a framework demonstrating how individuals living in low SES condition s have a smaller bank of resource s to cope with stressful events . According to this framework, limited coping resources lead to poorer health outcomes . According to these authors , e levated negative emotions and cognitions lead to psychological pathways causing poor long term health. Furthermore, i ndividual attitude s of mothers, which may include attitude s toward healthful eating (Clarke, Freeland Graves, Klohe Leman & Bohman, 2007; Jordan et al., 2008), mood self efficacy (Chang, Brown & Baumann, 2011), health motivation (Jayanti & Burns , 1998) and motivation to exercise, exercise self efficacy and overall self efficacy (Clarke et al., 2007; Mailey & McAuley , 2014 ), serve as predictive factors of maternal engagement in various health promoting lif estyle behaviors. Matthews, Gallo and Taylor (2010) also identify various positive psychological resources, such as self esteem and mastery, in the associatio n betwee n SES and health, but recognize a need to conduct further research on positive emotions th at lead to intermediate, mediating psychological pathways and health.

PAGE 19

9 According to Social Cognitive Theory (SCT ; Bandura, 1986 ) cognitive/personal (self efficacy) and social/environmental variables ( demographic and socioeconomic variab l es ) are important determinants of health behavior. SCT is often used to understand the underlying causes , or mechanisms, in the occurrence of health promoting behaviors (Tucker, Butler, Loyuk, Desmond & Surrency, 2009). Health Self Empowerment (HSE) Theory , a theo ry that builds upon SCT , may help to further explain differences in health promoting behaviors among members of population s who are dealing with the same social/environmental barriers that SCT cannot account for . HSE theory acknowledges the influence of en vironmental stressors, such as poverty, on health behaviors but asserts that key psychologica l variables and individual attitudes in minority populations and low income communities can significantly influence health behaviors. The current review of literature finds only one study which has examined the relationship between HSE t heory , including multiple and specific psychological variables and health promoting lifestyle behaviors. Tucker et al. (2009) examined this relationship among low income African American mothers and white mothers of chronically ill children. Active coping, health self efficacy, and health motivation were among the psychological variables that were found to significantly affect healthy diet, exerci s ing , using stress management practices , and engaging in health responsibility behaviors ( Abusabha & Achterberg , 1997; Tucker et al., 2009). Additionally, because a key facet of engaging in health promoting lifestyle behaviors is a pattern of self initiated actions and perceptions about individual health behavior and health wellness (Walker et al., 1987), l ocus of control may also be a psychological mechanism through which women engage in specific lifestyle behaviors . Rotter (1954) developed th e idea of individuals having either internal or external locus of control . Those with higher internal locus of

PAGE 20

10 control view events as resulting from their own actions, while those on the other end of the continuum, with higher external locus of control view events as being under the control of external factors. A multidimensional health locus of co ntrol construct has been used that assesses the specific role of beliefs in the context of health behavior, health outcomes and heal th care , and can help to contribute to the understanding of how perceived self efficacy and locus of control may be involved in the prediction of lifestyle behaviors ( Abusabha et al., 1997 ). Maternal depression is negatively correlated with health behaviors (Haller, Knisely, Dawson & Schnoll, 1993; Lindgren, 2001; W alker, Cooney, & Riggs, 1999). Thus, the presence of maternal depression is hypothesized to lead to a reduced likelihood of healthy behaviors . In the current study, we measure d maternal depression . In addition , we focused on one cognitive aspect of depression: learned helplessness . Learned helplessness is a related concept to locus of Hiroto & Seligman, 1975) the central idea is that if people feel they have no control over future outcomes, they are less likely to see k solutions to their problems. This chronic discovery of a loss of control leads to the concept of learned helplessness, a cognitive component of depression , and may be another mechanism that affects the relationship between environmental factors and lifestyle behaviors (Seligman, 1974) . Furthermore, these concepts are integrated in , from which HSE theory stems, in that self efficacy influences the expectations people hold about their abilities to control future outcomes and accomplish goals. Active coping and health motivation may positively correlate with self efficacy and internal locus of control and negatively correlate with learned helplessness. The current study measure d task oriented coping , health self efficacy, health motivation, and locus of control as psychological variables that may cause women to engage in health promotin g behaviors despite environmental, social and economic stress . U sing primar i ly

PAGE 21

11 HSE , the current study build s a conceptual framework (see Figure 1) to help show this framework . T he current study is the first of my knowledge that investigates multiple mediating variables that may influence the relationship between demographic and socioeconomic factors on lifestyle behaviors in a sample of mothers as a means of assessing greater hea lth disparity implications. Research Hypotheses The current study assess ed a multidimensional approach of socio demographic and socioeconomic factors with one approach being through the use of the Economic Hardship Scale ( EHS; Barrera et al., 2001). The EHS is a subjective measure of economic hardship that provides information about the psychological sense of disparity between needs and resources . A nother subjective indicator of socioeconomic position, the MacArth u r Scale o f Su bjective Social Status, was used to assess ocial status that reflects the relative perception that individuals have of their place in the social hierarchy . Subjective aspects of social position contribute to our understand i ng of the relationship between socioeconomic risk factors and health problems ( Wilkinson, 1997 ) as socioeconomic gradient and subjective experience of social status of individuals may be a better predict or of health than objective variables of social status , such as education or ma terial wealth (Everhart & Emde , 2006 ) . Each measure being used help s to understand unique dimensions of subjective life experience, including demographic aspects (access to health resources ), perceptions of social status, social support, and psychological sense of hardship. T he current study identified three separate models that represent this multidimensional approach . The current study examine d psychological variables that serve as potenti al mediators of the relationship between demograp hic and socioeconomic factors , and maternal engagement

PAGE 22

12 in health promoting lifestyle behaviors. Figure 1 depicts a visual representation of t he conceptual framework proposed in the study. Model 1 T he first model of the study examine d the influence of a subjective perspective of economic hardship on lifestyle behaviors. The current study is the first to my knowledge to examine the influence of a subjective perspective of socio demographic and socioeconomic fac tors on maternal engagement in health promoting lifestyle behaviors. Model 2 Secondly, because there is an abundance of evidence that socioeconomic status, educational achievement and social support significantly affect lifestyle behaviors and overall maternal and child health, the secondary model of the study was to measure these three variables. The second model of the study also meas ure d individual perception of social status to further understand another dimension of perception of hardship. Model 3 Las tly, access to resources was a third model of the study since research shows that the above independent variables, such as SES and ed ucational achievement , may affect receipt and access to health resources and care. It was expected that the relationship between socioeconomic r isk factors and engagement in health promo ting lifestyle behaviors would be significantly influenced by psychological mediating factors of mothers who come from underserved backgrounds. I predict ed that despite coming from low income, high stress environments and being exposed to socioeconomic risk factors, some mothers, because of individual psychologic al variables , such as health motivation, locus of control and self efficacy, will engage in health promoting lifestyle behaviors.

PAGE 23

13 CHAPTER 2 Method Participants Sample selection and eligibility criteria. Mothers were recruited from an advertisement to do an online surve y . The advertisement describe d the project as a study background and environment, lifestyle behaviors and individual attitude s. Participants wer e not provided with any monetary compensation for their participation . Informed consent was obtained from a brief initial section through the online survey. Design Th is study feature d a cross sectional design with a mediation analysis . The independent variable was defined as demographic and socioeconomic factors and include d : three socio demographic variables ( income , educational achievement level and level of access to resources), a total score for the experience of economic hardship , a tota l score for perceived social support, and two total score s for subjective social status (current and future perceptions) . The dependent variable was defined as maternal engagement in health promoting lifestyle behaviors and include s a total score of health promoting lifestyle behaviors. The mediating variable s were defined as the underlying psychological variables that are hypothesized to influence mothers from low income backgrounds to engage in health promoting behaviors and include health self efficacy, health motivation, task oriented coping , locus of control and depression . These variables and their measurement are described in greater detail in the

PAGE 24

14 Procedures Participants included 1 59 mothers who were asked to fill out a Qualtrics online survey through a link that was ey took approximately 30 40 minutes to complete and include d measures in a randomized order . Randomization was conducted by presenting qu estion blocks in random order using the Randomizer feature on Qu a ltrics. The scales listed b elow were converted to an online format which may affect the reliability of each scale. Reliability testing w as done to assess any changes in internal reliabilities of the scales. With the exception of the total score for access to health resources (.59) and the Subjective Social Status Scale (.68) .70 for all of the scales converted into online form and used in the current study. Participants had the option to leave their name and email and/or phone number if they were interested in participating in future studies . Measures Demographic Questionnaire. In order to assess objective measures of demographic and socioeconomic facto rs, the current study included a demographic questi onnaire. The questionnaire ask ed abou t age, ethnicity, race, marital/ relationship statu s, personal and family income, educ ational achievement and employment status, and access to medical and health resources . Economic Hardship . In order to assess a multidimensional perspective of socioeconomic and sociodemographic risk factors , the current study used the Economic Hardship Scale created by Barrera et al. (2001) . The Economic Hardship Scale is a 20 item self r eport

PAGE 25

15 measure and is a rated on a 5 The construct defines economic hardship as (a) the inability to afford specific necessities for living, (b) a general sense that financial or generate more income, and (d) hopelessness that the future will bring a brighter financial outlook. The construct shows r eliability in assessment of perceived economic hardship among multiple ethnic groups including African American, European American , and Mexican American families who reside in urban areas (Barrera et al., 2001 ) . Results of reliability analysis from the cur rent study are in line with previous research. provided for four subsets includes in the scale: Economic Adjustments and Cutbacks ( .79 ) Not Enough Money for Necessities ( .95 ), Inability to Make Ends Meet ( .86 ) and Financial Strain ( .80 ) disagree with these statements: We had enough money to afford the kind of car we need; In the past three months, we changed food shopping or eating This measure can objectively be obtained through a demographic questionnaire but also provides a subjective sense of hardship, individual need and personal and financial struggle. Subjective Social Status . The MacArthu r Scale of Subjective Social Status was administered as a measure of perceived social status across different levels of socioeconomic status (Adler & Stewart, 2007) . This measurement features a pictorial , and on which they feel they stand currently . Reliability testing has shown good internal reliability in previous research (Giatti, Valle Camelo, Rodriquez, & Barreta, 2012) , as well as in the current study . The current study also asked

PAGE 26

16 , between the two questions, current and future . This scale assesses common sense of social status across different SES and can include perceptions of community, familial and societal social statuses. Social Support. The Multidimensional Scale of Perceived Social Support is a 12 item self report measure of subjectively assessed social support , and is rated on a 5 point Likert type scale ranging from strongly disagree to strongly agree. The scale contains a total score 91 ), Friends (. 9 5) and Significant Other (.91). . Previous research has also shown good internal rel iability (Zimet, Powell, Farley, Werkman, & Berkoff, 1990). This measurement allowed us to assess how much they receive outside social support that can include family, friends and significant others. Active Coping. The current s tudy used the Coping Inventory for Stressful Situations Short Form (CISS SF; Cosway , Endler, Sadler, & Deary, 2000 ) which is a 21 item self report inventory created to measure task avoidance . This scale uses a 5 Not at All Very Much three subsets of the scale are: task oriented (. 84 ), emotional (. 84 ) and avoidant (. 65 ) , and (.74) for the total score . The CISS SF scale allowed us to assess coping styles of mothers and day to day coping o f stressful situations that play a role in physical and psychological well being.

PAGE 27

17 Health Self Efficacy. The School Health Efficacy Questionnaire (SHEQ) is a 43 item inventory that measures health self efficacy that is, the degree of confidence individuals have in their ability to perform specific health behaviors that contribute to their overall physical and mental well being. Most health behaviors measured by this scale are not specific to school, making it possible to use this scale with adults with mino r revisions . The current study used 35 out of 43 questions that applied to the target population, taking out questions regarding school health behaviors. The SHEQ has a seventh grade reading level, allowing measurement in groups of individuals from low edu cational achievement backgrounds. SHEQ responses are reported using a 5 point Likert scale ranging from very little to quite a lot. Reliability testing show a 91 for the total scale and is broken down into two subscales: Physical H ; of .80 ) and Mental ; of .83 ) Reliability testing has also shown good internal consistency in previous research (Froman & Owen, 1991). Because of the population being studied, the level of understanding of the SHEQ helped us to understand the self efficacy perceptions that motivate hea lth promoting lifestyle behaviors in mothers. Health Motivation. The Value on Health Scale (VHS) is a 5 item self report measure designed to assess value on, preference for, or personal importance of several aspects of health: fitness of being in good ph ysical condition, a sense of energy or vigor, endurance or stamina, maintaining an appro priate weight, and resistance to illness. The measure uses a 4 point Likert scale ranging from not important at all to very important. How

PAGE 28

18 alpha reliability of . 84 and has good internal reliabili ty according to previous studies ( Costa, Jessor, & Donovan, 1989). This scale was used to better understand how values placed on health are related to greater involvement in health enhancing behaviors, and how psychosocial and behavioral characteristics ma y explain differences in health behaviors. Locus of Control. The current study used the Multidimensional Health Locus of Control Scale (MHLOC) to assess locus of control regarding health behaviors (Wallston, Wallston & DeVellis, 1978) . The MHLOC is an 18 item self report measure that assesses about what influences health using a 6 . The scale assess es three dimensions of locus of control s alpha for the total scale is .7 0 ) : (1) internal belief (. 75 ) : and behaviors , (. 74 ) : me nor my doctor have much influence on it , and (3) powerful others belief (. 73 ) : dependent on the competence of my doctor . Reliability analyses in previous research shows adequate internal reliability (Thompson, Butcher & Berensen, 1987). The MHLOC offers an understanding of locus of control in reference to perceived self efficacy and its involvement in health behaviors. Maternal Depression. The Edinburgh Postnatal Depression Scale (EPDS; Cox et al., 1987) was used to assess symptoms of maternal depression. The EPDS is a 10 item self rating scale designed to detect postnatal depression and uses a 4 , . The total possible score ranges from 0 to 30 and indic ates the possible degree of postnatal depression exp erienced during the past week. Higher scores indicate more severe maternal depression. P revious research has found the best cutoff scores to be between 9.5 and 12.5 (Guedeney &

PAGE 29

19 Fermanian , 1998) , and multi ple studies have used a cutoff score of 9/10 to indicate the incidence of severe depressive symptoms (Jadresic, Araya & Jara, 1995) . The current study used the PEDS as a continuous score from 0 30, with higher values representing higher rates of depressive symptomology. The EPDS has been used with a variety of sample s and has shown good psychometric properties (Dayan et al., 2006; Eberhard Gran, Eskild, Tambs, Opjordsmoen & Samuelsen, 2001 ). Reliability studies confirm good internal consistency of the EPDS in postnatal depression symptomology evaluat ion (Bergink et al., 2011; Pop, Komproe & van Son, 1992; Teissedre & Chabrol, 2004) . C is .89 in the current sample. een so unhappy that I have been Learned Helplessness. Five questions from the Learned Helplessness Scale (LHS; Quinless & Nelson, 1988) were used to assess learned helplessness and combined with questions from the EPDS. The original LHS scale is a 20 item, self rating Likert type scale ranging from The total possible score of the original scale ranges from 20 to 80, with higher scores suggesting greater helplessness due to the perception that s control. The five questions that were used in t he current general If I complete a task successfully, it general , a Other people have more control over their success and/ or failure . Reliability 7 of the shortened version of the LHS

PAGE 30

20 used in the current study. Reliability analyses on the 20 item LHS also have confirmed good internal consistency (Quinless & Nelson, 1988) . Lifestyle Behaviors. The Health Promoting Lifestyle Behaviors Profile (HPLP II) was used as a multidimensional assessment of health promoting behavior patterns that include self initiated act promoting beh a viors. The HPLP II is a 52 item behavior rating scale that assesses the frequency of self reported health promoting behaviors in the domains of health responsibility, physical activity, nutrition, spirituality, interpersonal relations and stress management. There are six dimensions to the HPLP II scale: Self Actualization, Health Responsibility, Exercise, Nutrition, Internal Support and Stress Management . T he reliability is .9 5 . Reliability analyses confirm good internal consistencies in previous research (Walker, Noble, Sechrist, Richert, & Pender, 1987). saturated fa t and cholesterol , I follow a planned exercise program , relaxation each day , Data Analyses Primary Analysis I hypothesize d that the independent variable s of economic hardship, subjective social status, income, educational achievement and access to resources , and social support influence the mediator variable s , psychological variables including health self efficacy, health motivation, task oriented coping and health internal health locus of control , which in turn influence the dependent variable, health promoting lifestyle behaviors. In the statistical analysis, we tested this using a multiple mediation analysis considering our hypotheses that multiple mechanisms (med iators) are acting on this phenomena at once . The cu rrent study first used a serial multiple

PAGE 31

21 mediator model regression analysis , assuming the mediators are linked together in a causal chain (Hayes, 2013) . In a serial multiple mediator model, the total effe ct of X on Y partitions into direct and indirect components. The analysis includes obtaining t he direct effec t c 1 , w hich is interpreted as the estimated difference in Y between two cases that differ by one unit on X but are equal on all mediators. The serial multi ple mediator model contains indirect effects estimated as products of regression coefficients linking X to Y. The indirect effects, of which there are many due to the number of mediators in the model, are constructed by multiplying the weights corresponding to each step in an indirect pathway. These indirect effects are found in the PROCESS output along with 90 % bias corrected bootstrap confidence intervals based on 5,000 bootstrap samples. Both a 95% and a 90% confidence interval were used in t he analyses and no significant differences were found between the two. The numbers reported are using a 90% confidence interval in order to have a larger margin of error. I ndirect effects are all interpreted as the estimated difference in Y between two cas es that differ by one unit on X through the causal seq uence from X to mediators to Y. In order to express indirect effects in terms of the difference in standard deviations in Y between two cases that differ by one standard deviation in X, a completely sta ndardized effect size was used and denoted as c cs . Before testing the multiple mediators in the model, th e first step in the analysis is to confirm that at least two or more of the proposed mediators are correlated with each other, making the model a seria l multiple mediator model ( Hayes, 2013 ). Due to the significant correlations between the mediators (shown in Table 3 ), we used a serial multiple mediator model, or Model 6, in the PROCESS Macro. Since the serial multiple mediator model only allows for four mediators in the model at one time, health rel ated psychological mediators ( internal health locus of control, task oriented coping, health

PAGE 32

22 self efficacy and health motivation ) were run in the first step of analyses, while maternal depression and learned h elplessness were run in the second step of analyses. Secondary Analysis A secondary analysis was also done using structural equation modeling (SEM) to test the hypothesized relationships between demographic and socioeconomic factors, maternal lifestyle beh aviors and maternal depression, one of the proposed mediators. SEM has been described as a combination of confirmatory factory analysis and multiple regression (Ullman, 2001). SEM explores the possibility of relationships among latent variables and encompa sses two components: (a) a measurement model and (b) a structural model that allows the ability to test a full theoretical model among variables of interest that can be latent factors or directly measured. The measurement model of SEM is the confirmatory f actor analysis and depicts the pattern of observed variables for those latent constructs in the hypothesized model. The structural model comprises the other component of linear structural modeling and displays the interrelations among latent constructs and observable variables in the proposed model as a succession of structural equations. I performed an SEM analysis based on data from 288 mothers using the statistical modeling program Mplus 8 (Muthen & Muthen, 1994). Maximum likelihood (ML) parameter estim ation was chosen over other estimation methods (weighted least squares, two stage least squares) because the data were distributed normally, and ML estimation gives unbiased parameter estimates and standard errors for any missing data in the dataset (Schre iber, Nora, Stage, Barlow, & King, 2006). SEM estimated data using full maximum likelihood parameter estimation for 169 missing data points on the measure of maternal lifestyle behaviors (HPLP II).

PAGE 33

23 Three models were tested using Mplus to test three hypoth eses. The first model tested the first hypothesis of the theoretical dimensional structure of a latent construct, named personal income, family income, subjective economic har dship measured by the Economic Hardship Scale (EHS; Barrera et al.,) and subjective social status measured by the McArthur Scale of Subjective Social Status (SSS; Adler & Stewart, 2007) significantly comprise the underlying factor structure. The first mode l also tested the second hypothesis that the socioeconomic resources latent factor predicts engagement in maternal health behaviors, measured by the Health Promotion Lifestyle Profile (HPLP; Walker, Noble, Sechrist, Richert, & Pender, 1987). In the second and third models, the third hypothesis was tested and used SEM to explore the association of proposed mediators on the relationship between socioeconomic resources and maternal engagement in lifestyle behaviors. In the second model SEM was conducted to te st whether a proposed mediator, maternal depression, assessed using the Edinburgh Postnatal Depression Scale (EPDS; Cox et al., 1987), mediates the relationship between the socioeconomic resources latent factor and engagement in maternal health behaviors. Lastly, in the third model, SEM was conducted to test a post hoc hypothesis of whether two binary variables, years of education (high: 12 years or more of education; low: less than 12 years of education) and race (White; non White), significantly mediate t he relationship between socioeconomic resources and maternal lifestyle behaviors.

PAGE 34

24 CHAPTER THREE Results Sample Characteristics All participants had at least one child at or under the age of 10 years (exclusionary criteria if they did not). More than half of the sample (55 . 7 %) were between the ages of 25 and 3 4 years. Thirty four percent of the sample were betw een the ages of 35 and 44, 8 . 9 % of the sample were between the ages of 15 and 24 and 1.3% of the sample were between the ages of 45 an d 55 . Eighty four percent of mothers were of White or Caucasian decent, 11 % of mothers were of Hispani c or Latino decent, less than 1 % were of African American or Bl ack decent, less than 2 % were of Asian America n or Pacific Islander decent, about 3 % were o f multiracial decent and none were of American Indian or Alaskan Native decent . The majority of our sample had 12 or more years of education (83% ), and were employed (80.5% ). 56.6% of the sample ha d a personal income level of $39, 999 or less while 43.4% had a personal income of $40,000 or more. Additionally, 47.5% of the sample had a combined family income of $79,000 or less and 52.5% of the sample had a combined family income of $80,000 or more . The majority of the sample were eit her married or in a significant romantic relationship (89.9%). From the sample of 159 mothers, 102 mothers completed the entire survey, 46 did not fill out one or more full scales and 11 completed the survey with the exception of a few single missing times . For participants that were missing single items, data was imputed based on the average of other items in the same scale. Only one of the scales, the composite score for the Economic Hardship Scale (EHS) had a non normal distribution, with a skewness of 1 . 26 and kurtosis of 1. 01 . Association among Study Variables

PAGE 35

25 Primary Analysis Model 1 The first model of the study examined the influence of economic hardship on maternal health promoting lifestyle behaviors. M odel 1, using internal health locus of control, health motivation, health self efficacy and task oriented coping as mediators, is positive but not statistically significant , .01 , t (1 07 ) = . 3 1 , p = . 7 55 . Table 4 and figure 2 show s ignificant indirect effects found in this model i nclude the relationships between: 1) economic hardship, internal health locus of control, health motivation and lifestyle be h aviors was significant ( cs denotes a completely standardized effect size, cs = .01; Hayes, 2013 ) , 2 ) economic hardship, internal health locus of control, health self efficacy and lifestyle behavio rs wa s also significant ( cs = .01 ) , 3 ) economic hardship, internal health locus of control, task oriented coping and lifestyle behaviors was a lso significant ( cs = .01 ) and l astly, 4 ) economic hardship, internal health locus of control, health motivation, health self efficacy and lifestyle behaviors was signifi cant with an effect size of .00 2. Secondly, table 1 2 and figure 10 shows the direct effect of a separate analysis in model 1 , that us ed maternal depression and learned helplessness as mediators, was also positive but not statistically significant , 07 ) = . 3 12 p = . 7 55 ; there were no significant indirect effects found . Model 2 The second model of the study measure d socioeconomic status, both personal and familial, and educational achievement on maternal health promoting lifestyle behaviors . A separate analysis in model 2 also measure d individual perception of social status to further unde rstand another dimension of perception of hardship. Table 5 and figure 3 show the direct effect o f model 2 , with educational achievement as the independent variable, a nd internal health

PAGE 36

26 locus of control, health motivation, health self efficacy and task ori ented coping as mediators , was positive but not statistically significant , .0 6 , t (1 07 ) = 1. 37 , p = . 176 . The indirect effect s of the relationship between years of education, health motivation, health self efficacy, task oriented coping and lifestyle behaviors ( cs = .001), as well as years of education, task oriented coping and lifestyle behaviors were significant ( cs = .0 6 ) . Table 1 3 and figure 11 show the direct effect of a separate analysis in model two , using maternal depr ession and learned helplessness as mediators, was negative but not statistically significant , .06 , t (1 07 ) = 1. 37 , p = . 176 . Table 6 and figure 4 show t he direct effect of model 2 , with annual personal income as the independent variable, and internal health locus of control, health motivation, health self efficacy and task oriented coping as mediators , was positive but not statistically significant , . 04 , t (1 07 ) = 1. 96 , p = . 0 53 ; signifi cant indirect effects included the relationships bet ween 1) personal income, internal health locus of control, health motivation and lifestyle behaviors ( cs = .0 1 ), 2) personal income, internal locus of control, h ealth self efficacy and lifestyle behaviors ( cs = .0 1 ), 3 ) personal income, internal health locus of control, task oriented coping, and lifestyle behaviors ( cs = .01), 4 ) personal income, internal health locus of control, health motivation, health self efficacy and lifestyle behaviors ( cs = .00 3 ), 5 ) personal income, internal health locus of control, health self efficacy, task oriented coping, and lifestyle behaviors ( cs = .00 04 ) , and lastly, 6 ) personal income, health motivation, health self efficacy , task oriented coping and lifestyle behaviors ( cs = .00 04 ) . Table 1 4 and figure 12 show t he direct effect of model 2 , with annual personal income as the independent variable, and maternal depression and learned helplessness as mediators, was positive .0 4 , t (10 7 ) = 1. 96 , p = . 0 53 .

PAGE 37

27 Table 7 and figure 5 show t he direct effect of model 2 with annual family income as the independent variable, and internal health locus of control, health motivation, health self efficacy and task oriented coping as mediators , was positive but no t statistically significant , 2 , t (1 07 ) = . 8 7 , p = . 385 . Th e indirect effect of the relationship between family income, health motivation, health self efficacy, task oriented coping and lifestyle behaviors was significant ( cs = .001) . Table 1 5 and figure 13 show t he direct effect of model 2 with annual family income as the independent variable, and maternal depression and learned helplessness as mediators, was 2 , t (1 07 ) = . 8 7 , p = . 385 ; there were no significant indirect effects . Table 8 and figure 6 show t he direct effect of model 2 with social support as the independent variable, and internal health locus of control, health motivation, health self efficacy and task oriented coping as mediators , (1 07 ) = . 63 , p = . 5 30 . The indirect effect s of the relationship between social support, internal locus of control, health self efficacy, task oriented coping and lifestyle behaviors ( cs = .001) and social support, internal health locus of control, health motivation, health self efficacy, task oriented coping and lifestyle behaviors ( cs = .0002) were statistically significant . Table 1 6 and figure 14 show t he direct effect of model 2 with social support as the independent variable, and maternal depression and learned helplessness as mediators, was also positive but not statistically 07 ) = . 62 , p = . 5 30 ; there were no significant indirect effects. Table 9 and figure 7 show t he direct effect of model 2 with subjective social status, current, as the independent variable, and internal health locus of control, health motivation, health self efficacy and task oriented coping as mediators , was negative but not stati stically .0 1 , t (1 07 ) = . 56 , p = . 577 ; there were no significant indirect effects . Table 1 7

PAGE 38

28 and figure 15 show t he direct effect of model 2 with subjective social status, current, as the independent variable, and maternal depression and learned helplessness as mediators, was also negative .0 1 , t (1 07 ) = . 56 , p = . 577 with no significant direct effects . Table 10 and figure 8 show t he direct e ffect of model 2 with subjective social status, future, as the independent variable, and internal health locus of control, health motivation, health self efficacy and task oriented coping as mediators , was positive but not statistically significant, 01, t (1 07 ) = . 23 , p = . 8 23 . There were four indirect effects found to be significant in this model: 1) the indirect effect between subjective social status, future, internal health locus of control, health motivation , health self efficacy and lifestyle behaviors was significant ( cs = .0 0 2) 2) indirect effect between subjective social status, future, internal health locu s of control, health self efficacy, task oriented coping and lifestyle behaviors was significant ( cs = .0 0 1 ) ; 3) the indirect effec t between subjective social status, internal health locus of control, health motivation, heath self efficacy, task oriented coping and healthy lifestyle behaviors was significant ( cs = .00 03 ); 4) the indirect effect between subjective social status, futu re, health motivation, health self efficacy, task oriented coping and lifestyle behaviors was significant ( cs = .001 ). T able 1 8 and figure 16 show the direct effect of model 2 with subjective social status, future, as the independent variable, and maternal depression and learned helplessness as 07 ) = . 23 , p = . 8 23 . Model 3 Thirdly, access to health resources was a third model of the study . Table 1 1 and figure 9 show t he direct e ffect of model 3 with access to health resources as the independent variable, and internal health locus of control, task oriented coping, health self efficacy and health motivation

PAGE 39

29 4 , t ( 1 07 ) = . 9 6 , p = . 3 38 . There were s ix significant indirect effects found in this model: 1) access to healthcare resources, internal health locus of control, health motivation and lifestyle behaviors ( cs = .0 2 ) 2 ) the relationship between access to healthcare resources, internal health locus of control, health self efficacy and lifestyle behaviors ( cs = .01), 3 ) the relationship between access to healthcare resources, internal health locus of control, task oriented coping, and lifestyle behaviors ( cs = .01), 4 ) the relationship between access to healthcare resources, internal health locus of control, health motivation, health self efficacy and lifestyle behaviors ( cs = .003), 5 ) the relationship between access to healthcare resources, internal health locus of control, health self efficacy, task oriented coping and lifestyle behaviors ( cs = .00 2 ), 6 ) the relationship between access to healthcare resources, internal health locus of control, health motivation, health self efficacy, task oriented coping and lifestyle behaviors (.001) . Table 1 9 and figure 17 show t he direct effect of model 3 with access to health resources as the independent variable, and maternal depression and learned helplessness as mediators, was also positive but not statistica 4 , t (1 07 ) = . 9 6 , p = . 3 38 with no significant indirect effects. Secondary Analysis We ran a structural equation model to test the hypotheses that 1) a socioeconomic resources variable would have good factor structure using personal income, family income, subjective economic hardship and subjective social status, 2) that the socioeconomic resources latent variable would significantly predict maternal engagement in health promoting lifestyle behaviors and 3) that maternal depression, years of education and race would also significantly mediate this relationship. First, fit statistics of the first model ( figure 22 ) confirmed that the data fit the model for each component of socioeconomic resources and the prediction of the latent

PAGE 40

30 constr uct of socioeconomic resources on maternal engagement in health promoting lifestyle behaviors, Chi Squared (5, N = 288) = 10.76, p = .056, root mean square error of approximation (RMSEA) = .06, comparative fit index (CFI) = .97. Fit statistics of the secon d model (f igure 2 3 ) confirmed adequate fit of maternal depression mediating the relationship between the latent construct of socioeconomic resources and maternal engagement in health promoting lifestyle behaviors, Chi Squared (8, N = 288) = 22.44, p = .004 , RMSEA = .08, CFI = .94, with a non significant standardized indirect effect of .01, p = .423. Fit statistics for the third model ( figure 24 ) confirmed good fit of years of education and race mediating the relationship between the latent construct of socioeconomic resources and maternal engagement in health promoting lifestyle behaviors, Chi Squared (12, 288) = 17.53, p = .131, RMSEA = .04, CFI = .98, with no indirect effects reported in the model due to the binary variables used. Post Hoc Analyses Economic Hardship Subscale scores Post hoc analyses were conducted using the four subsets of the economic hardship scale as the independent variables instead of the economic hardship composite score that was used in the main analyses. Table 20 and figure 1 8 show the direct effect of the post hoc analysis with inability to make ends meet as the independent variable and internal health locus of control, task oriented coping, health self efficacy and health motivation as mediators, was positive but not statis Table 21 and figure 1 9 show the direct effect of the post hoc analysis with adjustments and cutbacks as the independent variable and internal health locus of con trol, task oriented coping, health self efficacy and health motivation as mediators, was also positive but not statistically Table 22 and figure 20 show the direct effect of the

PAGE 41

31 post hoc analysis with not e nough money for necessities as the independent variable and internal health locus of control, task oriented coping, health self efficacy and health motivation as mediators, was .22, t (107) = .2.84, p = .006. Finally, t able 20 and figure 20 show the direct effect of the post hoc analysis with financial strain as the independent variable and internal health locus of control, task oriented coping, health self efficacy and health motivation as mediators, was posi .02, t (107) = .19, p = .847.

PAGE 42

32 CHAPTER FOUR Discussion The purpose of the current study was to examine the influence of psychological variables on the relationship between demographic and socioeconomic variables and maternal lifestyle behaviors . As a means to further understand the proximate causes that create health disparities , the current study hypothesized that psychological mediators would signific antly influence the relationship between sociodemographic and socioeconomic factors and maternal lifestyle behaviors . I did not find any significant direct mediation effects of psychological variables on the relationship between sociodemographic and socioe conomic variables and maternal lifestyle behaviors in the current sample. The current study identified three separate models to: 1) examine the influence of a subjective perspective of economic hardship on lifestyle behaviors through the influenc e of psych ological mediators, 2) assess the relationship between socioeconomic status, educational achievement and social support on maternal lifestyle behaviors through the influence of psychological mediators and 3) assess the relationship between maternal access to health resources, through the influence of psychological mediators, on maternal health behaviors . The results showed that despite insignificant findings of the overall mediating effects of psychological variables on the relationship between specific demographic and socioeconomic factors and maternal lifestyle behaviors , there are statistically significant indirect pathways between the variables that were investigated . The results of the current study show tentative indirect relationships between spec ific psychological mediators and maternal lifest yle behaviors. R elative to those with lower subjective experience of hardship, those with a higher subjective perspective of hardship had a lower engagement in health promoting lifestyle behaviors through the pathways of lower internal

PAGE 43

33 health locus of control, lower health motivation, health self efficacy and task oriented coping style . Higher educational achievement resulted significant indirect pathways of higher health motivation, health self efficacy, and task oriented coping, leading to higher engagement in health promoting lifestyle behaviors. Additionally, those with higher annual personal income had higher engagement in health promoting lifestyle behaviors through the pathways of internal health locus o f control, health motivation, health self efficacy and task oriented coping. Family income also had one significant indirect effect showing higher family income resulting in higher engagement in health promoting lifestyle behaviors through the pathways of higher health motivation , task oriented coping, and health self effi cacy. Social support played a similar role, with higher social support leading to higher engagement in health promoting lifestyle behaviors through the pathways of higher internal health locus of control , health motivation, health self efficacy, and task o riented coping. An additional measure of subjective perception of hardship, anticipated future subjective social status, was significantly indirectly and positively related to maternal health behaviors through multiple pathways, including internal health l ocus of control, task oriented coping, health motivation, and heath self efficacy . Lastly, higher access to health resources also resulted in higher engagement in health promoting lifestyle behaviors through pathways of higher internal health locus of cont rol, health motivation, health self efficacy and task oriented coping. Results of three models of SEM in the secondary analysis conclude that 1) the underlying factor structure of the latent construct fit the data well and the latent construct of socioeco nomic resources significantly predicts maternal enga gement in lifestyle behaviors, 2 ) maternal depression does not significantly mediate the relationship between the latent construct socioeconomic resources and maternal engagement in health pro moting lifes tyle behaviors and

PAGE 44

34 3 ) years of education and race do not significantly mediate the relationship between the latent construct socioeconomic resources and maternal engagement in health promoting lifestyle behaviors. Post hoc analyses also found that one subs et of the EHS significantly predicted engagement in health prom oting lifestyle behaviors. The n ot enough money for necessities subset score significantly negatively predicted engagement in health promoting lifestyle behaviors, showing this subset of the EHS may capture a monetary predictor of engagement in kind of food we We capture a better depiction of monetary resources that result in an increase or decrease in engagement in healthy lifestyle behaviors. However, this mod el displays a significant direct effect of not enough money for necessities and maternal engagement in health promoting lifestyle behaviors; a full mediation is not supported since the independent variable does not significantly predict the psychological m ediators. One potential explanation for why the current study did not find significant direct main effects may be due to the demographics of the sample. The recruitment method was unable capture a diverse sample of mothers from both high and low socioecon omic backgrounds, ethnically and racially diverse backgrounds and of a diverse group of educational and occupational experiences. Future research may examine the direct relationships between the demographic and socioeconomic variables and psychological med iators , and maternal lifestyle behaviors . It may be that the various demographic and socioeconomic variables significantly predict maternal lifestyle behaviors independently, as shown in previous research ( Matijasevich et al., 2010; Premji, 2014 ) . P sychological mediators may also independently predict maternal

PAGE 45

35 engagement in lifestyle behaviors. It may be helpful to investigate the relationship between psychological mediators and maternal lifestyle behaviors while controlling for demographic and socio economic variables. Future research should primarily establish these two relationships prior to conducting a mediational analysis combining the target variables. To date, the current study is the first to look at multiple mediating variables that may influ ence the relationship between sociodemographic and socioeconomic risk factors on lifestyle behaviors in a sample of mother s. Although direct model findings were insignificant, the current study tentatively provides information on indirect pathways through wh ich psychological mediators may influence maternal lifestyle behaviors. Implications The purpose of the current study was to examine the relationship s between demographic and socioeco nomic factors and maternal engagement in health promoting lifestyle be haviors. Fur thermore, the current study aimed to investigate whether individual attitudes, or specific ps ychological variables , influence the above relatio nship. The current study did not provide support for the HSE theory in understanding health promoting lifestyle behaviors and l evels of engagement. Specifically, the aim of the study was to extend the findings of previous research that shows these self empowerment or self motivation behaviors are particularly influential o n lifestyle behaviors and choices. Furthermore, our mediation hypothesis was not supported . However, the significant indirect effects found may help to identify tentative mechanistic pathways that explain why some women engage in health promoting lifestyle behaviors despite dealing with socioeconomic and sociodemographic stressors. Previous research in self motivation attitudes has found that specific factors may influence attitudes. Chang et al. (2011) found that availability of time mothers had to prepare

PAGE 46

36 food affected their engagement in health promoting dietary behaviors, and a higher positive mood increased self efficacy and healthy choices. Additionally, maternal outcome of expectancies may also affect maternal engagement in healthy lifestyle choices. Chang et al. (2008) also found that if mothers are aware of the beneficial effects of healthy choices, they are more likely to engage in health promoting behaviors despite dealing with challenges associated with low income or underserved backgrounds. Socia l support may also impact attitude s toward behaviors and the success of behavioral interventions, as Clarke et al. (2007) found positive social support to be a significant contributor to engagement in healthy lifestyle choices and behavior, specifically in mothers . The findings of the current study suggest that variables such as self efficacy, health motivation, task oriented coping and loc us of control are predictive of maternal health behaviors (b pathways), but do not significantly influence the relation ship between the predictor and outcome variables. Furthermore, it may be that the proposed mediators serve as independent variables and covariates along with the other proposed independent variables in the current study. This is in line with the previous l iterature mentioned above, as well as a possible explanation of the null findings. Instead of serving as the mechanisms through which demographic and socioeconomic factors influence maternal lifestyle behaviors, it may make more sense to view psychological mediators as influences on maternal lifestyle behaviors on their own. If future studies are able to replicate this model with a population of specifically low income mothers, it may be that the proposed mediators do influence the relationship between demographic and socioeconomic factors and maternal lifestyle behaviors. Thr ough the increased engagement in health promoting lifestyle behaviors, the current study w as able to show how to

PAGE 47

37 potentially m aximize health promoting lifestyle behaviors in mothers, through the influence of psychological mediators. This may be able to hel p healthcare providers, social workers, psychologists, teachers and other services workers within low income communities to assess for and fosters health promotion focused lifestyle choices and behaviors through the use of these psychologi cal mediators. Te aching skills in health motivation and health self efficacy, for example, through health behavior interventions , may potentially prevent serious and chronic health problems, such as cardiovascular disease, diabetes, and a variety of psychological disorders , in a variety of populations by increasing engagement of health promoting lifestyle behaviors (Walsh, 2011). Modeling of maternal engagement in health promoting lifestyle behaviors in low income and underserved communities may also have even greater impli cations in health outcomes by teaching children and younger generations about health promoting lifestyle behaviors and choices which may eventually decrease health disparities in this population. Study Limitations The fin dings of the current study need t o be interpreted with caution based on some limitations of the study . First, aim was to recruit a diverse sample of mothers from underserved backgrounds. A lthough successful in recruiting a diverse group of mothers in some demographics, such as age , recruitment via Facebook was not successful in reaching mothers from significantly underserved backgrounds (e.g., annual income of <$10,000) . Additionally, based on the demographic results of the analysis, our sample primarily consisted of Ca ucasian women (83.6%), therefore, external validity is threaten ed . The mechanistic pathways that were conditionally found in a somewhat homogeneous sample may be more clearly defi ned in a more diverse sample that includes more women from lower income backgrounds. The use of

PAGE 48

38 Facebook for recruitment threatened external validity and generalizability of results and created a selection bias that affected the independent variables by the voluntary and self selected recruitment method (e.g., demographic and socioeconomic factors and the mediators, self efficacy, locus of control, etc.). Additionally, convenience sample from Facebook recruitment skew ed the data in that we were unable to capture women who do not have Facebook or access to technology and social media. Thirdly, the use of a cross sectional research design is a significant limitation of the study. An important point of m ediation analyses is that the independent variable must precede the dependent variable, which inherently implies causation, with mediator variables thought to be causal variants between the criterion and outcome variables. Previous research has investigated whether mediation can be tested with cross sectional data. Some researchers have concluded that the res ults of analyses based on cross sectiona l data are u nlikely to accurately reflect longitudinal mediation effects , and that using longitud inal designs ena bles researchers to make more r igorous inferences about causal relations implied by mediation models (C ole & Maxwell, 2003). The results of the current study may reflect a bias or misleading estimate of the relationship between the target variables. If socioeconomic and demographic variables, psychological mediators and maternal lifestyle behaviors had been studied at two or three different time points, different mediation results may have been shown . Potential threats to the internal and statistical conclusion validity of the study also deserve mention . C orrections of controlling for inflation and type I error were planned in the analysis and methods sections by splitting up the indepen dent variables into three separate analyses and using a Hayes multiple mediation analysis to decrease the number of regressions run and minimize increase d inflation o f type one and two errors for each analysis being done . A

PAGE 49

39 bias corrected confidence interval of 90% was also used. However, it is still possible that we failed to detect significant results because the current study did not have adequate power with the cur rent sample size relative to the number of variables used . Additionally, because the scales being used in the online survey have been formally used and validated in paper format, this factor may also pose a threat to internal validity. Psychometric tests ( e.g., assessing for to assess validity and reliability w ere performed during analysis to assess for this potential threat to validity. One scale was particularly low in relia health resources items was .59, indicating a significantly lower reliability between the scale items. However, although there were some consequences of changing the mode of administration for the surveys being used (e.g., difficult to control study environment, unmonitored study participants, and selection bias) , there have been found to be benefits of online research studies. Greater population access, less experimental costs, no time limitations and complete voluntary participation which may indicate participant motivation (Musch & Reips , 2000) are examples of beneficial differences that were found to make internet based data collection a suitable alternative to in person administration, keeping in mind the d ata collection medium creates a suitable but not identical alternative ( Riva, Teruzzi, & Anolli, 2003) .

PAGE 50

40 REFERENCES Abel, M. H. (1996). Inadequate Prenatal Care. Psychological Reports, 79, 903 912. Adler, N., & Stewart, J. (2007). The MacArthur scale of subjective social status. MacArthur Research Network on SES & Health. Retrieved from http://www. macses. ucsf. edu/Research/Psychosocial/subjective. php . Abusabha, R., & Achterberg, C. (1997). Review of self efficacy and locus of control for nutrition and health related behavior. Journal of the American Dietetic Association , 97 (10), 1122 1132. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory . Prentice Hall, Inc. Baron, R.M., & Kenny, D. A. (1986). The moderator mediat or variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of personality and social psychology, 51(6), 1173. Barrera, M., Caples, H., and Tein, J. (2001). The psychological sense of economic ha rdship: Measurement models, validity, and cross ethnic equivalence for urban families. American Journal of Community Psychology , 29 (3), 493 517. Bergink, V., Kooistra, L., Lambregtse van den Berg, M. P., Wijnen, H., Bunevicius, R., van Baar, A., & Pop, V. (2011). Validation of the Edinburgh Depression Scale during pregnancy. Journal of psychosomatic research , 70 (4), 385 389. Biddle, S. J., & Asare, M. (2011). Physical activity and mental health in children and adolescents: a review of reviews. British Journa l of Sports Medicine , 45 (11), 886 95. http://doi.org/10.1136/bjsports 2011 090185 Blair, C., Granger, D. A., Willoughby, M., Mills Fortunato, C. K. (2011). Salivary Cortisol Mediates Effects of Poverty and Parenting on Executive Functions in Early Childhood. Child Development , 82 (6), 1970 1984. http://doi.org/10.1111/j.1467 8624.2011.01643.x Buford, T. W., Anton, S. D., Clark, D. J., Higgins, T. J., & Cooke, M. B. (2014). Optimizing the benefits of exercise on physic al function in older adults. Function, and Rehabilitation , 6 (6), 528 43. http://doi.org/10.1016/j.pmrj.2013.11.009 Chang, M. W., Nitzke, S., Brown, R., & Baumann, L. (2011). Predictors of Low income, Obese althful Weight Management Behaviors. Journal of Nutrition Education and Behavior , 43 (2), 87 95. http://doi.org/10.1016/j.jneb.2009.11.006 Clarke, K. K., Freeland Graves, J., Klohe Lehman, D. M., & Bohman, T. M. (2007). Predictors of Weight Loss in Low Inco me Mothers of Young Children. Journal of the American Dietetic Association , 107 (7), 1146 1154. http://doi.org/10.1016/j.jada.2007.04.016

PAGE 51

41 Clearfield, M. W., Carter Rodriguez, A., Merali, A. R., & Shober, R. (2014). The effects of SES on infant and maternal diurnal salivary cortisol output. Infant Behavior and Development , 37 (3), 298 304. http://doi.org/10.1016/j.infbeh.2014.04.008 Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with longitudinal data: questions and tips in the use of structural equation modeling. Jo urnal of abnormal psychology , 112 (4), 558. Corwin. (2008). Immune dysregulation and glucocorticoid resistance in minority and low income pregnancy women. October 141 (4), 520 529. http://doi.org/10.1016/j.surg.2006.10.010.Use Costa, F. M., Jessor, R., & Don ovan, J. E. (1989). Value on health and adolescent conventionality: A construct validation of a new measure in problem behavior theory. Journal of Applied Social Psychology , 19 (10), 841 861. http://doi.org/10.1111/j.1559 1816.1989.tb01225.x Cosway, R., End ler, N. S., Sadler, A. J., & Deary, I. J. (2000). The coping inventory for stressful situations: Factorial structure and associations with personality traits and psychological health 1 . Journal of Applied Bio Behavioral Research , 5 (2), 121 143. Cox, J. L., Holden, J. M., & Sagovsky, R. (1987). Detection of postnatal depression. Development of the 10 item Edinburgh Postnatal Depression Scale. The British journal of psychiatry , 150 (6), 782 786. Dayan, J., Creveuil, C., Marks, M. N., Conroy, S., Herlicoviez, M ., Dreyfus, M., & Tordjman, S. (2006). Prenatal depression, prenatal anxiety, and spontaneous preterm birth: a prospective cohort study among women with early and regular care. Psychosomatic Medicine , 68 (6), 938 946. Dennis, J. (2013). Young maternal age a nd low birth weight risk: An exploration of racial/ethnic disparities in the birth outcomes of mothers in the United States, 141 (4), 520 529. http://doi.org/10.1016/j.surg.2006.10.010. Dog, L. (2010). The role of nutrition in mental health. Alternative therapies in health and medicine , 16 (2), 42. Dole, N., Savitz, D. A., Herz Picciotto, I., Siega Riz, A. M., McMahon, M. J., & Buekens, P. (2003). Maternal Stress and Preterm Birth. American Journal of Epidemiology , 157 (1), 14 24. http://doi.org/10.1093/aje /kwf176 Dominguez, T. P. (2011). Adverse birth outcomes in African American women: the social context of persistent reproductive disadvantage. Social Work in Public Health , 26 (1), 3 16. http://doi.org/10.1080/10911350902986880

PAGE 52

42 Duffy, M. E. (1997). Determin ants of reported health promotion behaviors in employed Mexican American women. Health Care for Women International , 18 (2), 149 163. http://doi.org/10.1080/07399339709516270 Eberhard Gran, M., Eskild, A., Tambs, K., Opjordsmoen, S., & Ove Samuelsen, S. (20 01). Review of validation studies of the Edinburgh Postnatal Depression Scale. Acta Psychiatrica Scandinavica , 104 (4), 243 249. Essex, M., Klein, M., Cho, E., & Kalin, N. H. (2002). Maternal stress beginning in infancy may sensitize children to later stres s exposure: Effects on cortisol and behavior. Biological Psychiatry, 52(8), 776 784. Evans, G. (2003). A multimethodological analysis of cumulative risk and allostatic load among rural children. Developmental Psychology, 39(5), 924 933 . http://dx.doi.org/1 0.1037/0012 1649.39.5.924 Everhart, K., & Emde, R. N. (2006). Perspectives on stress and self regulatory processes. The Crisis in Youth Mental Health: Childhood disorders , 1 , 1. Fink, N. S., Urech, C., Cavelti, M., & Alder, J. (2012). Relaxation During Pre gnancy. The Journal of Perinatal & Neonatal Nursing , 26 (4), 296 306. http://doi.org/10.1097/JPN.0b013e31823f565b Fritz, M. S., & MacKinnon, D. P. (2007). Required sample size to detect the mediated effect. Psychological science , 18 (3), 233 239. Froman, R. D., & Owen, S. V. (1991). High school students' perceived self efficacy in physical and mental health. Journal of Adolescent Research , 6 (2), 181 196. Fiscella, K., & Williams, D. R. (2004). Health disparities based on socioeconomic inequities: implications for urban health care. Academic Medicine: Journal of the Association of American Medical Colleges , 79 (12), 1139 1147. http://doi.org/10.1097/00001888 200412000 00004 Gazmararian, J. A., Adams, M. M., & Pamuk, E. R. (1996). Associations between measures of socioeconomic status and maternal health behavior. American Journal of Preventive Medicine , 12 (2), 108 15. Retrieved from http://www.ncbi.nlm.nih.gov /pubmed/8777063 Giatti, L., Camelo, L. D. V., Rodrigues, J. F. D. C., & Barreto, S. M. (2012). Reliability of the MacArthur scale of subjective social status Brazilian Longitudinal Study of Adult Health (ELSA Brasil). BMC Public Health , 12 (1), 1096. http ://doi.org/10.1186/1471 2458 12 1096 Gennaro, S. (2003). Psychological and Physiological Stress: Impact on Preterm Birth. Journal of Obstetric, Gynecologic, and Neonatal Nursing , 32 (5), 668 675. http://doi.org/10.1177/0884217503257484

PAGE 53

43 Ghosh, J. K. C., Wilhelm, M. H., Dunkel Schetter, C., Lombardi, C. A., & Ritz, B. R. (2010). Paternal support and preterm birth, and the moderation of effects of chronic stress: A study in Los Angeles County mothers. , 13 (4 ), 327 338. http://doi.org/10.1007/s00737 009 0135 9 Gortmaker, S. L. (1979). The effects of prenatal care upon the health of the newborn. American Journal of Public Health , 69 (7), 653 660. http://doi.org/10.2105/AJPH.69.7.653 Grossman, P., Niemann, L., Sc hmidt, S., & Walach, H. (2004). Mindfulness based stress reduction and health benefits: A meta analysis. Journal of Psychosomatic Research , 57 (1), 35 43. http://doi.org/10.1016/S0022 3999(03)00573 7 Guedeney, N., & Fermanian, J. (1998). Validation study of the French version of the Edinburgh Postnatal Depression Scale (EPDS): new results about use and psychometric properties. European psychiatry , 13 (2), 83 89. Haller, D. L., Knisely, J. S., Dawson, K. S., & Schnoll, S. H. (1993). Perinatal substance abuser s: Psychological and social characteristics. The Journal of Nervous and Mental Disease , 181 (8), 509 513. Hardie, J. H., & Landale, N. S. (2013). Profiles of Risk: Maternal Health, Socioeconomic Status, and Child Health. Journal of Marriage and Family , 75 (3 ), 651 666. http://doi.org/10.1111/jomf.12021 nets. Journal of Marriage and Family, 73 (4), 861 875. http://doi.org/10.1038/nature13314.A Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression based approach . Guilford Press. W. (2014). Socioeconomic status and depression during and after pregnancy in the Franconian Maternal Health Evaluation Studies (FRAMES). Archives of Gynecology and Obstetrics , 289 (4), 755 63. http://doi.org/10.1007/s00404 013 3046 y Hiroto, D. S. & Seligma n, M.E. (1975). Generality of learned helplessness in man. Journal of Personality and Social Psychology, Vol 31(2 ), 311 327. http://dx.doi.org/10.1037/h0076270 Hobel, C. J., Dunkel Schetter, C., Roesch, S . C., Castro, L. C., & Arora, C. P. (1999). Maternal plasma corticotrophin pregnancies ending in preterm delivery. American journal of obstetrics and gynecology , 180 (1), S257 S263.

PAGE 54

44 Jadresic , E., Araya, R., & Jara, C. (1995). Validation of the Edinburgh postnatal depression scale (EPDS) in Chilean postpartum women. Journal of Psychosomatic Obstetrics & Gynecology , 16 (4), 187 191. James, S. A. (1993). Racial and ethnic differences in infant mo rtality and low birth weight. A psychosocial critique. Annals of Epidemiology , 3 , 130 136. http://doi.org/http://dx.doi.org/10.1016/1047 2797(93)90125 N Jayanti, R. K., & Burns, A. C. (1998). The Antecedents of Preventive Health Care Behavior: An Empirical Study. Journal of the Academy of Marketing Science , 26 (1), 6 15. http://doi.org/10.1177/0092070398261002 Jordan, K. C., Freeland Graves, J. H., Klohe Lehman, D. M., Cai, G., Voruganti, V. S., Proffitt, al activity intervention promotes weight loss and enhances diet attitudes in low income mothers of young children. Nutrition Research , 28 (1), 13 20. http://doi.org/10.1016/j.nutres.2007.11.005 Kahn, H. S., Williamson, D. F., & Stevens, J. A. (1991). Race a nd weight change in US women: The roles of socioeconomic and marital status. American Journal of Public Health , 81 (3), 319 323. http://doi.org/10.2105/AJPH.81.3.319 Lancaster, C. A., Gold, K. J., Flynn, H. A., Yoo, H., Marcus, S. M., & Davis, M. M. (2010). Systematic Review, 202 (1), 5 14. http://doi.org/10.1016/j.ajog.2009.09.007.Risk Lantz, Paula M., et al. "Socioeconomic disparities in health change in a longitudinal study of US adults: the role of health risk behaviors." Social science & medicine 53.1 (2 001): 29 40. Luecken, L. J. (2008). Prenatal Stress, Partner Support, and Infant Cortisol Reactivity in Low Income Mexican American Families. October 141 (4), 520 529. http://doi.org/10.1016/j.surg.2 006.10.010 . Luecken, L. J., Lin, B., Coburn, S. S., MacKinnon, D. P., Gonzales, N. A., & Crnic, K. A. (2013). Prenatal stress, partner support, and infant cortisol reactivity in low income Mexican American families. Psychoneuroendocrinology , 38 (12), 3092 3101. Mailey, E. L., & McAuley, E. (2014). Impact of a brief intervention on physical activity and social cognitive determinants among working mothers: A randomized trial. Journal of Behavioral Medicine , 37 (2), 343 355. http://doi.org/10.1007/s10865 0 13 9492 y Matijasevich, A., Victora, C. G., Lawlor, D. A., Golding, J., Menezes, A. M. B., Araújo, C. L., Barros, A. J., Santos, I. S., Barros, F. C., & Smith, G. D. (2012). Association of socioeconomic position with maternal pregnancy and infant health ou tcomes in birth cohort studies from Brazil and the UK. Journal of Epidemiology and Community Health , 66 (2), 127 35. http://doi.org/10.1136/jech.2010.108605

PAGE 55

45 Matthews, K. A., Gallo, L. C., & Taylor, S. E. (2010). Are psychosocial factors mediators of socioec onomic status and health connections? Annals of the New York academy of sciences , 1186 (1), 146 173. McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiological Reviews , 87 , 873 904. http://doi.org/10 .1152/physrev.00041.2006 Low early life social class leaves a biological residue manifested by decreased glucocorticoid and increased proinflammatory signaling. Proceedings of the Nat ional Academy of Sciences of the United States of America , 106 (34), 14716 21. http://doi.org/10.1073/pnas.0902971106 H., & Grason, H. A. (2002). Associations between maternal and child health status and patterns o f medical care use. Ambulatory Pediatrics , 2 (2), 85 92. http://doi.org/10.1367/1539 4409(2002)002<0085:abmach>2.0.co;2 Mitchell, T. M. D. F., & Barlow, C. E. M. S. (2011). Review of the Role of Exercise in Improving Quality of Life in Healthy Individuals a nd in Those with Chronic Diseases. Current Sports Medicine Reports , 10 (4), 211 216. Retrieved from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&CSC=Y&NEWS=N&PAGE=fulltext&D=ovftl &AN=00149619 201107000 00011 Musch, J., & Reips, U. D. (2000). A brief history of W eb experimenting. Retrieved from psychnet.apa.org Parker, J. D., Schoendorf, K. C., & Kiely, J. L. (1994). Associations between measures of socioeconomic status and low birth weight, small for gestational age, and premature delivery in the United States. A nnals of Epidemiology , 4 (4), 271 278. http://doi.org/10.1016/1047 2797(94)90082 5 Prather, H., Spitznagle, T., & Hunt, D. (2012). Benefits of exercise during pregnancy. PM and R , 4 (11), 845 850. http://doi.org/10.1016/j.pmrj.2012.07.012 Premji, S. (2014). Perinatal Distress in Women in Low and Middle Income Countries: Allostatic Load as a Framework to Examine the Effect of Perinatal Distress on Preterm Birth and Infant Health. Maternal and Child Health Journal , 18 (10), 2393 2407. http://doi.org/10.1007/s10 995 014 1479 y Pop, V. J., Komproe, I. H., & Van Son, M. J. (1992). Characteristics of the Edinburgh post natal depression scale in The Netherlands. Journal of affective disorders , 26 (2), 105 110. Program for prevention research. (1999). Manual for the chi and the how I coped under pressure scales. Available from Arizona State University, P.O. Box 876005, Tempe, AZ 85287 6005 , 7420 (480).

PAGE 56

46 Quinless, F. W., & Nelson, M. A. M. (1988). Development of a measure of learned helple ssness. Nursing Research , 37 (1), 11 15. Riva, G., Teruzzi, T., & Anolli, L. (2003). The use of the internet in psychological research: comparison of online and offline questionnaires. CyberPsychology & Behavior , 6 (1), 73 80. Rotter, J. B. (1954). Social learning and clinical psychology. Englewood Cliffs: Prentice Hall. Roux, A. V. D. (2013). Conceptual Approaches to the Study of Health Disparities, 41 58. http://doi.org/10 .1146/annurev publhealth 031811 124534.Conceptual Santiago, C. D., Wadsworth, M. E., & Stump, J. (2011). Socioeconomic status, neighborhood disadvantage, and poverty related stress: Prospective effects on psychological syndromes among diverse low income fa milies. Journal of Economic Psychology , 32 (2), 218 230. http://doi.org/10.1016/j.joep.2009.10.008 Schaefer, J. T., & Magnuson, A. B. (2014). A review of interventions that promote eating by internal cues. Journal of the Academy of Nutrition and Dietetics , 114 (5), 734 760. http://doi.org/10.1016/j.jand.2013.12.024 Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J. (2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. The Journal of educational research , 99 (6), 323 338. Seligman, M. E. (1974). Depression and learned helplessness. In R. J. Friedman & M. M. Katz (Eds.), The psychology of depression: Contemporary theory and research. Oxford, England: John Wiley. Song, J. E., Kim, T., & Ahn, J. A. (2015). A Systematic Review of Psychosocial Interventions for Women with Postpartum Stress. JOGNN Journal of Obstetric, Gynecologic, and Neonatal Nursing , 44 (2), 183 192. http://doi.org/10.1111/1552 6909.12541 Teissèdre, F., & Chabrol, H. (2004). Detecting women at risk for postnatal depression using the Edinburgh Postnatal Depression Scale at 2 to 3 days postpartum. Canadian Journal of Psychiatry , 49 (1), 51 54. Thayer, Z. M., & Kuzawa, C. W. (2014). Early origins of health disparities: Material deprivation predicts maternal evening cortisol in pregnancy and offspring cortisol reactivity in the first few weeks of life. American Journal of Human Biology , 26 (6), 723 730. http://doi.org/10.1002/ajhb.22532 Thoits, P. A. (2010). Stress and Health: Major Findings and Policy Implications. Journal of Health and Social Behavior , 51 (1 Suppl), S41 S53. http://doi.org/10.1177/0022146510383499

PAGE 57

47 Thompson, B., Butcher, A., & Beren reliability and validity study. Measurement and Evaluation in Counseling and Development, 20, 80 88 . Tucker, B. (2016). Predictors of a Health Promoting Lifestyle and Behaviors Among Low Income African American Mothers and White Mothers of Chronically Ill Children. Ullman, J. B. (2001). Structural equation modeling. In B. G. Tabachnick & L. S. Fidell (Eds.), Using multivariate statistics (4th ed.). Needham Heights, MA: Allyn & Bacon. Uriu Adams, J. Y., Obican, S. G., & Keen, C. L. (2013). Vitamin D and maternal and child health: Overview and implications for dietary requirements. Birth Defects Research Part C Embryo Today: Reviews , 99 (1), 24 44. http://doi.org/10.1002/bdrc.21031 Wallston, K. A., Wallston, B. S., & DeVellis, R. (1978). Development of the multidimensional health locus of control (MHLC) scales. Health Education & Behavior , 6 (1), 160 170. Walker, L. O., Cooney, A. T., & Riggs, M. W. (1999). Psychosocial and demographic factors relat ed to health behaviors in the 1st trimester. Journal of Obstetric, Gynecologic, & Neonatal Nursing , 28 (6), 606 614. Walker, S. N., Sechrist, K. R., & Pender, N. J. (1987). The Health Promoting Lifestyle Profile. Nursing Research, 36 (2). doi:10.1097/00006199 198703000 00002 Walsh, R. (2011). Lifestyle and Mental. Journals@Ovid Full TextAmerican Psychologist , 66 (7), 579 592. http://doi.org/10.1037/a0021769 Weitz, T. A., Freund, K. M., & Wright, L. (2001). Identifying and caring for underserved populations: experience of the National Centers of Excellence in Women's Health. Journal of women's health & gender based medicine , 10 (10), 937 952. Wilkinson, R. G. (1997). Socioeconomic determinan ts of health. Health inequalities: relative or absolute material standards? BMJ (Clinical Research Ed.) , 314 (7080), 591 595. http://doi.org/10.1136/bmj.314.7080.591 Zimet, Powell, S., Farley, G., Werk man, S., & Berkoff, K. (1990). Psychometric characteristics of the Multidimensional Scale of Perceived Social Support. Journal of Personality Assessment , 55 (3 4), 610 617. http://doi.org/10.1080/00223891.1990.9674095

PAGE 58

48 APPENDIX Table 1 Demographics of Study Sample Age Category N (%) 15 24 14 (8.8) 25 34 88 (55.3) 35 44 54 (34.0) 45 55 2 (1.3) Ethnicity Category N (%) African American or Black 1 (.6) White or Caucasian 133 (83.6) Latino or Hispanic 17 (10.7) Asian American or Pacific Islander 3 (1.9) Multiracial 5 (3.1) % Married 89.9 % % Completed HS or above 83 % % Unemployed 16.4 %

PAGE 59

49 Table 2 Descriptive statistics for all variables used in mediation analyses Variable M SD N Access to Healthcare Resources 1.09 1.22 159 Economic Hardship Scale (EHS) composite score . 18 3.5 6 1 25 Inability to Make Ends Meet subset total score 2.44 1. 12 125 Not Enough Money for Necessities subset total score 2.13 1.00 125 Adjustments and Cutbacks subset total count 1.52 1.97 125 Financial Strain subset total score 1.34 .72 125 Subjective Social Status current (SSS1) 5.60 1.83 12 5 Subjective Social Status future (SSS2) 6.38 1.79 125 Multidimensional Scale of Perceived Social Support (MSPSS) 5. 91 1. 44 129 Edinburg Postnatal Depression Scale (EPDS) 8. 27 5. 34 127 Learned Helplessness Scale (LHS) 8.86 3.04 127 Coping Inventory for Stressful Situations ( CISS; total) 64.16 10.35 127 Coping Inventory for Stressful Situations ( CISS; task oriented) 24.97 5.40 127 Health Self Efficacy Questionnaire ( SHEQ; mental health) 57.60 7.63 132 Health Self Efficacy Questionnaire ( SHEQ; physical health) 81.74 8.60 132 Health Self Efficacy Questionnaire ( SHEQ; total) 147.74 16.10 132 Value on Health Scale (VHS) 15.43 3.10 134 Multidimensional Health Locus of Control ( MHLOC; internal ) 31.28 5.34 126 Multidimensional Health Locus of Control ( MHLOC; chance) 20.06 6. 27 126 Multidimensional Health Locus of Control ( MHLOC; powerful others) 18.81 6.26 126 Health Promoting Lifestyle Behaviors ( HPLP; total) 2.63 .50 123

PAGE 60

50 Table 3 Correlations of all variables Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1. Age __ .12 .46** .48** .03 .003 .35** .05 .07 .27** .05 .03 .03 .14 .04 .06 .10 .14 .13 .05 .10 .02 .25** .01 2. Employ ment Status .12 __ .48** .08 .12 .07 .06 .12 .06 .16 .09 .01 .04 .10 .00 .02 .10 .04 .06 .04 .11 .14 .00 .04 3. Personal Income .46** .48** __ .57** .01 .10 .15 .33** .27** .27** .23* .17 .12 .34** .12 .21* .28** .26** .28** .25** .17 .04 .01 .30** 4. Family Income .48** .08 .57** __ .05 .12 .32** .42** .53** .21* .18* .14 .10 .31** .06 .02 .26** .31* .31** .20* .01 .04 .09 .25** 5. Years of Education .03 .12 .01 .05 __ .05 .004 .12 .19* .07 .07 .004 .04 .01 .06 .20* .11 .12 .13 .11 .04 .10 .19* .06 6. Relationship Status .003 .07 .10 .12 .05 __ .001 .01 .20* .04 .10 .21* .04 .06 .04 .09 .01 .04 .0 2 .01 .09 .13 .05 .10 7. Race/Ethnicity .35** .06 .15 .32 .004 .001 __ .15 .10 .08 .05 .001 .05 .10 .04 .11 .11 .22* .18* .09 .13 .11 .19* .23

PAGE 61

51 8. Access to Healthcare (higher scores equal less access) .05 .12 .33 .42* .12 .01 .15 __ .69** .06 .08 .32** .34** .25** .02 .17 .48** .44** .49** .23** .01 .03 .01 .27** 9. EHS composite .07 .06 .27** .53** .19** .20** .10 .69** __ .16 .11 .36** .46** .43** .07 .21* .52** .48** .53** .24** .11 .12 .02 .36** 10. SSS1 .27** .16 .27** .21** .07 .04 .08 .06 .16 __ .52** .02 .11 .10 .004 .08 .02 .07 .05 .1 2 .04 .07 .07 .02 11. SSS2 .05 .09 .23* .18* .07 .10 .05 .08 .11 .52** __ . 13 .00 .05 .12 .12 .13 .09 .12 .19* .08 .02 .04 .20* 12. MSPSS .03 .01 .17 .14 .004 .21* .001 .32** .36** .02 .13 __ .37** .30** .15 .28** .48** .40** .47** .21** .05 .13 .01 .42** 13. EPDS .03 .04 .12 .10 .04 .04 .05 .34* .46* .1 1 .00 .37** __ .61 .14 .23* .69** .62** .68** .17 .13 .29** .00 .49** 14. LHS .14 .10 .33** .31** .01 .06 .10 .25** .43** .10 .05 . 30** .61** __ .12 .19** .54** .53** .56** .21* 14 .39** .10 .42** 15. CISS total .04 .00 .12 .06 .06 .04 .04 .02 .07 .004 .12 .15 .14 .12 __ .61** .06 .06 .08 .01 .05 .04 .23* .28** 16. CISS task oriented .06 .02 .21* .02 .20* .09 .11 .17 .21* .08 .12 .28** .23* .19* .61** __ . 38** .33** .38** . 16 .26** .18 .13 .46** 17. SHEQ mental health .09 .10 .28** .26** .11 .01 .11 .48** .52** .02 .13 .48** .69** .54** .06 .38** __ .78** .93** .30** .18 .40** .003 .62**

PAGE 62

52 18. SHEQ physical health .14 .04 .26** .31 ** .12 .04 .22* .44** .48** .07 .09 .40** .62** .53** .06 .33** .78 ** __ .95** .36** .19* .33** .04 .65** 19. SHEQ total .13 .06 .28** .31** .13 .0 2 .18* .49** .53** .05 .12 .47** .68** .56** .08 .38** .93** .95** __ .34** .18* .37** .01 .68** 20. VHS .05 .04 .25** .20* .11 .01 .09 .23** .24** .12 .19* .21* .17 .21* .01 .16 .30** .36** .34** __ .37** .21** .06 .49** 21. MHLOC internal .10 .11 .17 .01 .04 .09 .13 .01 .11 .04 .08 .05 .13 .14 .05 .26** .18 .19* .18* .37** __ .38 .21 .34** 22. MHLOC chance .02 .14 .04 .04 .10 .13 .11 .03 .12 .07 .02 .13 .29** .39** .04 .18 .40** .33** .37** . 21** .38** __ .31** .33** 23. MHLOC powerful others .25** .00 .01 .09 .19* .05 .19* .01 .02 .07 .04 .01 .00 .10 .23* .13 .003 .04 .01 .06 .21* .31* __ .07 24. HPLP .01 .04 .28** .25** .06 .10 .23* .27** .36** .02 .20* .42** .49** .42** .28** .46** .62** .65** .68** .49** .39** .33** .07 __

PAGE 63

53 Table 4 Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi ) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff . SE p X (EHS) . 5 2 .27 .05 9 .02 .15 .9 14 .1 5 .51 .77 8 . 1 9 .26 .475 .01 .0 2 . 7 55 M1 (MHLOCi) __ __ __ .20 .0 6 < .001 *** .36 .2 1 .0 87 .21 .11 .047 * .01 .01 .141 M2 (VHS) __ __ __ __ __ __ . 5 7 .36 .120 .1 3 .19 .472 .03 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 9 .05 .1 18 .01 .00 .00 1 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .00 5 ** Constant 32.79 6.13 <.001 *** 6.007 3.74 .108 133.79 1 3.15 <.001** * 1 .31 9.72 .893 . 2 4 .6 1 . 702 R 2 = . 22 F ( 14, 92 ) = 1. 83 , p = . 046* R 2 = . 27 F ( 15, 91 ) = 2.27 , p = .00 9 ** R 2 = .67 F ( 16, 90 ) = 11.42 , p < .001 *** R 2 = . 30 F ( 17, 89 ) = 2. 28 , p = .007 * * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001 *** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e H PLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, * **correlation is significant at <.001 level.

PAGE 64

54 Table 5 Regression of educational achievement on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE P Coeff. SE p X (Education) .06 .75 .934 .3 8 .39 .345 1. 91 1.39 .171 1.0 6 .71 .138 .06 .05 .176 M1 (MHLOCi) __ __ __ .2 0 .0 6 <.001 * ** . 3 6 .2 1 .087 .21 .11 .047* .01 .01 .141 M2 (VHS) __ __ __ __ __ __ . 5 7 .36 .1 20 . 1 3 .19 .472 .0 3 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 8 .05 .118 .01 .00 .001 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .00 5 ** Constant 32.79 6.13 <.001** * 6.07 3.74 < .001 *** 133.79 13.11 < .001* ** 1 .31 9.72 .893 .24 .61 .702 R 2 = .22 F (1 4, 92 ) = 1. 83 , p = . 046* R 2 = .2 7 F (1 5, 91 ) = 2. 27 , p = .00 9 ** R 2 = . 67 F (1 6, 90 ) = 11.42 , p < .001* ** R 2 = . 30 F (1 7, 89 ) = 2. 28 , p = .007 * * R 2 = .6 5 F (1 8, 88 ) = 9.08 , p < .001** * Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Health Promoting Lif estyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel.

PAGE 65

55 Table 6 Regression of personal income on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Personal Income) .86 .34 .014* . 09 .19 .648 .6 5 .6 5 .320 . 54 .33 .109 .0 4 .0 2 . 0 53 M1 (MHLOCi) __ __ __ .20 .0 6 < .001** * .35 . 21 .0 87 .21 .11 .047 .01 .01 .141 M2 (VHS) __ __ __ __ __ __ . 5 7 .36 .120 . 1 3 .19 .472 .03 .01 .0 16 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 8 .05 .118 .01 .00 .001 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .005 * * Constant 32.79 6.13 <.001* ** 6.07 3.74 .108 133.79 13.11 < .001 *** 1 .31 9.71 .893 .24 .61 .702 R 2 = .22 F (1 4, 92 ) = 1. 83 , p = . 046* R 2 = .27 F (1 5, 91 ) = 2. 27 , p = .00 9 ** R 2 = . 67 F (1 6, 90 ) = 11.42 , p < .001 *** R 2 = . 3 F (1 7, 89 ) = 2. 28 , p = .007 * ( R 2 = .65 F (1 8, 88 ) = 9.09 , p < .001 *** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel

PAGE 66

56 Table 7 Regression of family income on maternal lifestyle behaviors wit h psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Family Income) .27 .42 .523 . 19 .22 .384 .1 8 .77 .8 19 . 19 .39 .6 25 .0 2 .0 3 . 385 M1 (MHLOCi) __ __ __ .2 0 .0 5 < .001 *** . 3 6 . 2 1 . 0 87 . 21 .1 1 . 047* .01 .01 .140 M2 (VHS) __ __ __ __ __ __ . 5 8 .36 .119 . 1 3 .1 9 .472 .03 .01 .0 16 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 9 .05 . 118 .01 .00 .001 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .005 * * Constant 32.79 6.13 <.001* ** 6.07 3.74 .108 133.79 11.42 < .001 *** 1 .31 9.71 . 893 .24 . 61 .702 R 2 = .22 F (1 4, 92 ) = 1. 83 , p = . 046* R 2 = .27 F (1 5, 91 ) = 2. 27 , p = .00 9 ** R 2 = .67 F (1 6, 90 ) = 11.42 , p < .001 *** R 2 = . 30 F (1 7, 89 ) = 2. 28 , p = .007* * R 2 = .6 5 F (1 8, 88 ) = 9.08 , p < .001 *** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel.

PAGE 67

57 Table 8 Regression of social support on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Social Support) .38 .54 .476 . 09 .29 .742 1.1 0 .99 .269 . 53 .5 0 .292 .0 2 .03 .5 30 M1 (MHLOCi) __ __ __ .2 0 .05 <.001 * ** . 3 6 . 21 .086 .21 .11 .047* .01 .01 .141 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .1 20 . 1 3 .19 . 472 .03 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ . 0 8 .0 5 .1 18 .01 .00 .00 1 * * M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .00 5 * * Constant 32.79 6.13 <.001* ** 6.07 3.74 .108 1 33.79 13.11 < .001* ** 1 .31 9.72 .893 .24 .61 .702 R 2 = .22 F ( 14, 92 ) = 1.83 , p = . 046* R 2 = .2 7 F ( 15, 91 ) = 2. 27 , p = .00 9 * * R 2 = .67 F ( 16, 90 ) = 11.42 , p < .001* ** R 2 = . 30 F ( 17, 89 ) = 2. 28 , p = .007* * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001 *** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e H PLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, * **correlation is significant at <.001 level.

PAGE 68

58 Table 9 Regression of Subjective Social Status (SSS1; current) on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (SSS1) . 53 .42 .2 03 .0 8 .2 2 .7 09 . 3 5 .77 .6 54 . 42 .39 .279 .0 1 .0 3 . 577 M1 (MHLOCi) __ __ __ .2 0 .0 6 <.001*** .36 .2 1 .0 87 . 21 .048* .123 .01 .01 .141 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .1 20 . 1 3 .18 .472 .03 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 8 .05 .1 18 .01 .00 .00 1 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .00 5 ** Constant 32.79 6.13 <.001 *** 6.07 3. 74 .108 1 33.79 11.42 < .001*** 1 .31 9.72 .893 .24 . 61 .702 R 2 = .22 F ( 14, 92 ) = 1.83 , p = . 046* R 2 = .2 7 F ( 15, 91 ) = 2. 27 , p = .00 9 ** R 2 = .67 F ( 16, 90 ) = 11.42 , p < .001*** R 2 = . 30 F ( 17, 89 ) = 2. 28 , p = .007 * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001*** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e H PLP = Health Promoting Lif estyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel.

PAGE 69

59 Table 10 Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with psycholog ical mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (SSS2) .45 .40 .266 . 18 .21 .402 . 5 8 .74 .434 .3 5 .38 .361 .0 1 .02 .8 23 M1 (MHLOCi) __ __ __ .2 0 .0 6 <.001*** . 3 6 . 21 .0 87 .2 1 .1 1 . 047* .01 .01 .141 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .119 . 1 3 .1 9 . 472 .03 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 8 .05 .1 18 .01 .00 .00 1 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .00 5 ** Constant 32.79 6.13 <.001 *** 6.07 3.74 .108 1 33.79 1 3.11 <.001*** 1 .31 9.71 .893 .24 .61 .702 R 2 = .22 F ( 14, 92 ) = 1.83 , p = . 046 R 2 = .2 7 F ( 15, 91 ) = 2.27 , p = .00 9 ** R 2 = .67 F ( 16, 90 ) = 11.42 , p < .001*** R 2 = . 30 F ( 17, 89 ) = 2.28 , p = .007* * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001*** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e H PLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel.

PAGE 70

60 Table 1 1 Regression of access to health resources on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Access to Healthcare) 1. 18 .66 .078 . 1 3 .36 .7 13 1. 7 6 1. 2 4 .15 9 . 37 .6 3 .5 59 .0 4 .04 .338 M1 (MHLOCi) __ __ __ . 2 0 .0 6 <.001** * . 36 . 21 .0 87 .21 .1 1 . 047 * .01 .01 .141 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .1 19 . 1 3 .1 9 . 472 .03 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 8 .05 . 118 .01 .00 .001 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .00 5 ** Constant 32.79 6.13 <.001* ** 6.07 3.74 .108 1 33.79 11.42 < .001* ** 1 .31 9.72 .893 .24 .61 .702 R 2 = .22 F ( 14, 92 ) = 1.83 , p = . 046* R 2 = .2 7 F ( 15, 91 ) = 2 .27 , p = .00 9 * * R 2 = .67 F ( 16, 90 ) = 11.42 , p < .001** * R 2 = . 30 F ( 17, 89 ) = 2. 28 , p = .007* * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001** * Note . a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e H PLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, * *correlation is significant at .01 level, ***correlation is significant at <.001 level

PAGE 71

61 Table 1 2 Regression of Economic Hardship Scale (EHS) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (EHS) . 4 7 . 19 .0 12 * . 05 .1 2 .69 4 .01 .0 2 . 7 55 M1 (EPDS) __ __ __ .1 9 .0 7 .00 6 ** .01 .01 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.31 5.67 < .001** * 18.25 3.99 <.001* ** .24 .6 1 . 702 R 2 = . 63 F ( 16, 90 ) = 9.67 , p < .001** * R 2 = .5 5 F ( 17, 89 ) = 6.40 , p < .001** * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Table 1 3 Regression of educational achievement on maternal lifestyle behaviors with maternal de pression and learned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X ( Yrs. Education) 1.0 4 .52 .0 47* . 1 3 .3 3 .6 95 .06 .05 . 178 M1 (EPDS) __ __ __ . 1 9 .0 4 .00 6 ** .01 .01 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5.67 < .001** * 18.25 3.99 < .001*** .23 . 61 .702 R 2 = .63 F ( 16, 90 ) = 9.67 , R 2 = .5 5 F ( 17, 89 ) = 6.81 , R 2 = .65 F ( 18, 88 ) = 9. 08 ,

PAGE 72

62 p < .001** * p < .001** * p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 l evel. Table 1 4 Regression of personal income on maternal lifestyle behaviors with maternal depression and learned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (Personal Income) . 20 .24 .412 .2 2 .15 .1 52 .0 4 .0 2 .0 53 M1 (EPDS) __ __ __ .1 9 .0 7 .00 6 ** .01 .0 1 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5.84 < .001** * 18.25 3.99 <.001* ** .24 .61 .702 R 2 = .6 3 F ( 16, 90 ) = 9.67 , p < .001** * R 2 = . 55 F ( 17, 89 ) = 6.40 , p < .001** * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Table 1 5 Regression of family income on maternal lifestyle behaviors with maternal depression and learned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (Family Income) . 2 7 .28 .3 28 .39 . 17 .026 .0 2 .0 3 .385

PAGE 73

63 M1 (EPDS) __ __ __ .1 9 .0 7 .00 6 ** .01 .01 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5.67 < .001** * 18.25 4.57 <.001* ** .24 .61 .702 R 2 = .6 3 F ( 16, 90 ) = 9.67 , p < .001** * R 2 = . 55 F ( 17, 89 ) = 6.40 , p < .001** * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Table 1 6 Regression of social support on maternal lifestyle behaviors with maternal depression and learne d helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (Social Support) . 6 1 .3 6 .101 . 3 5 .2 3 . 129 .02 .03 . 631 M1 (EPDS) __ __ __ .1 9 .0 6 .00 6 ** .01 .01 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5.67 < .001** * 18.25 3.99 <.001* ** .24 .61 .702 R 2 = .6 3 F ( 16, 90 ) = 9.67 , p < .001** * R 2 = . 55 F ( 17, 89 ) = 6.40 , p < .001** * R 2 = .65 F ( 18, 88 ) = 9. 08 , p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.

PAGE 74

64 Table 1 7 Regression of Subje ctive Social Status (SSS1; current) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (SSS1) .6 8 .2 8 .018 * .00 .18 .998 .0 1 .0 3 . 577 M1 (EPDS) __ __ __ .1 9 .0 7 .06 4 ** .01 .01 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5.67 < .001** * 18.25 3.99 <.001* ** .24 .6 1 . 702 R 2 = .6 3 F ( 16, 90 ) = 9.67 , p < .001** * R 2 = . 55 F ( 17, 89 ) = 6.40 , p < .001** R 2 = .65 F (1 8, 88 ) = 9. 08 , p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Table 1 8 Regression of Subjective Social Status (SSS2; future) on maternal lifestyle behaviors with maternal depression and learned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (SSS2) . 09 .28 .741 .0 5 .1 7 . 77 8 .01 .02 .8 23 M1 (EPDS) __ __ __ .1 9 .0 6 .00 6 ** .01 .01 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5.67 < .001** * 18.25 3.99 <.001* ** .24 .61 .702 R 2 = .63 F ( 16, 90 ) = 9.67 , R 2 = . 55 F ( 17, 89 ) = 6.40 , R 2 = .65 F ( 18, 88 ) = 9. 08 ,

PAGE 75

65 p < .001** * p < .001** * p < .001** * Note . a EPDS = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Table 1 9 Regression of access to health resources on maternal lifestyle behaviors with maternal depression and lear ned helplessness mediators Consequent Antecedent M1 (EPDS) M2 (LHS) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p X (Access to Healthcare) . 29 .46 .52 7 . 45 .2 9 .121 .0 4 .04 . 3 38 M1 (EPDS) __ __ __ . 1 9 .0 6 .00 6 ** .01 .01 . 336 M2 (LHS) __ __ __ __ __ __ .0 1 .0 2 . 626 Constant 29.32 5. 67 < .001 *** 18.25 3.99 <.001** * .24 .61 . 702 R 2 = .6 3 F ( 16, 90 ) = 9.67 , p < .001* * R 2 = . 55 F ( 17, 89 ) = 6.40 , p < .001** * R 2 = .64 F ( 18, 88 ) = 9. 08 , p < .001** * = Edinburgh Postnatal Depression Scale, b LHS = Learned helplessness scale (shortened), c HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.

PAGE 76

66 Table 20 Regression of inability to make ends meet subscale of EHS on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Inability to Make Ends Meet) . 15 1.30 .909 .0 1 .69 .9 91 2.00 2.39 .404 . 34 1.21 .777 .0 9 .0 8 .263 M1 (MHLOCi) __ __ __ .20 .05 <.001*** .3 6 .21 .086 .21 .11 .047 * .01 .01 . 131 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .1 20 .1 3 .19 .477 .03 .01 .015 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .0 9 .05 .128 .01 .00 .002 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .005 ** Constant 33.18 7. 04 <.001* ** 6.09 4.18 .149 1 28.50 14.57 < .001*** .58 10.10 .954 .42 .6 4 .512 R 2 = .22 F ( 15, 91 ) = 1.69 , p = .067 R 2 = .27 F ( 16, 90 ) = 2.11 , p = .014 * R 2 = .67 F ( 17, 89 ) = 10.76 , p < .001*** R 2 = .30 F ( 18, 88 ) = 2.13 , p = .011 * R 2 = .66 F ( 19, 87 ) = 8.69 , p < .001*** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, * *correlation is significant at .01 level, ***correlation is significant at <.001 level.

PAGE 77

67 Table 21 Regression of adjustments and cutbacks subset of EHS on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Adjustments and Cutbacks) . 20 .60 .736 .0 4 .32 .907 . 20 1.11 .858 . 84 . 55 .133 .0 4 .04 .227 M1 (MHLOCi) __ __ __ .2 1 .0 6 <.001*** .3 6 .21 .088 . 20 .1 1 .052 .01 .01 .138 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .1 22 .1 3 .1 8 .477 .03 .01 .016 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .09 .05 .10 9 .01 .00 .002** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .003 ** Constant 33.02 6.20 <.001* ** 6.02 3.78 .115 133.53 13.26 < .001*** 2.22 9.66 .819 .28 .61 .645 R 2 = .22 F ( 15, 91 ) = 1.70 , p = . 065 R 2 = .27 F ( 16, 90 ) = 2.11 , p = .014 * R 2 = .67 F ( 17, 89 ) = 10.64 , p < .001*** R 2 = .32 F ( 18, 88 ) = 2.31 , p = .005* * R 2 = .66 F ( 19, 87 ) = 8.73 , p < .001*** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Heal th Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel

PAGE 78

68 Table 22 Regression of not enough money for necessities on maternal lifestyle beha viors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X (Not Enough for Money Necessities) . 68 1.35 .618 . 44 .72 .539 1. 43 2.49 .569 1.69 1.25 .19 . 22 .08 .006** M1 (MHLOCi) __ __ __ . 20 .05 <.001*** .3 5 .21 .091 .21 .11 .044 * .01 .01 .160 M2 (VHS) __ __ __ __ __ __ .5 6 .36 .131 . 1 2 .19 .517 .03 .01 .017 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .09 .05 .099 .01 .00 .002** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .001 ** Constant 34.35 6.90 <.001* ** 7.15 4.14 .088 137.34 14.55 < .001*** 3.49 10.30 .735 . 38 .63 .544 R 2 = .22 F ( 15, 91 ) = 1.71 , p = . 063 R 2 = .27 F ( 16, 90 ) = 2.14 , p = .013 * R 2 = .67 F ( 17, 89 ) = 10.69 , p < .001*** R 2 = .32 F ( 18, 88 ) = 2.27 , p = .006* * R 2 = .68 F ( 18, 87 ) = 9.72 , p < .001*** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Health Promoting Lifestyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level

PAGE 79

69 Table 23 Regression of financial strain subset on maternal lifestyle behaviors with psychological mediators Consequent Antecedent M1 (MHLOCi) M2 (VHS) M3 (SHEQ) M4 (CISSto) Y (HPLP) Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p Coeff. SE p X ( Financial Strain) .8 8 1.29 .498 .2 3 .69 .737 .6 7 2.37 .778 . 09 1.20 .938 .0 2 .08 .847 M1 (MHLOCi) __ __ __ .20 .06 <.001*** .36 .21 .087 .21 .11 .049 * .01 .01 .147 M2 (VHS) __ __ __ __ __ __ .5 7 .36 .119 .11 .19 .56 4 .0 3 .01 .017 * M3 (SHEQ) __ __ __ __ __ __ __ __ __ .09 .05 .1 20 .01 .00 .001 ** M4 (CISSto) __ __ __ __ __ __ __ __ __ __ __ __ .02 .01 .005 ** Constant 31.70 6.36 <.001* ** 5.83 3.83 .131 134.48 13.39 < .001*** 1.20 9.88 .904 .25 .62 .686 R 2 = .22 F ( 15, 91 ) = 1.73 p = . 059 R 2 = .27 F ( 16, 90 ) = 2.12 , p = .0 14 * R 2 = .67 F ( 17, 89 ) = 10.65 , p < .001*** R 2 = .30 F ( 18, 88 ) = 2.13 , p = .011 * R 2 = .65 F ( 189, 87 ) = 8.51 , p < .001*** Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e HPLP = Health Promoting Lif estyle Profile * Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 le vel

PAGE 80

70 Model 1 Model 2 Model 3 Figure 1. Diagram of Conceptual framework of three models. Note. a MHLOCi = Multidimensional health locus of control, internal subscale, b VHS = Value on health scale, c SHEQ = School health self efficacy questionnaire, d CISSto = Coping inventory for stressful situations, task oriented subscale, e EPDS = Edinburgh Postnatal Depression scale, f LHS = Learned Helplessness Scale, g HPLP = He alth Promoting Lifestyle Profil Access to health resources MHLOCi VHS SHEQ CISSto EPDS LHS Analysis 1 Analysis 2 HPLP Income (personal and family) Education Social Support SSS1 SSS2 MHLOCi VHS SHEQ CISSto HPLP EPDS LHS Analysis 1 Analysis 2 EHS MHLOCi VHS SHEQ CISSto EPDS LHS Analysis 1 Analysis 2 HPLP

PAGE 81

71 Fig ure 2 . Diagram of psychological mediators on relationship between economic hardship and maternal lifestyle behaviors. Note . a X = Economic Hardship Scale, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correla tion is significant at <.001 level. Figure 3 . Diagram of psychological mediators on relationship between educational achievement and maternal lifestyle behaviors. Note . a X = Years of education , b M1 = Multidimensional health locus of control, internal su bscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation i s significant at .01 level, ***correlation is significant at <.001 level.

PAGE 82

72 Figure 4 . Diagram of psychological mediators on relationship between annual personal income and maternal lifestyle behaviors. Note . a X = Annual p ersonal income, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 5 . Diagram of psychological mediators on relationship between annual family income and maternal lifestyle behavior s. Note . a X = Annual f amily income, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile

PAGE 83

73 *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 6 . Diagram of psychological mediators on relationship between social support and maternal lifestyle behaviors. Note . a X = Multidimensional Scale of Perceived Social Support, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 7 . Diagram of psychological mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors. Note . a X = Subjective Social Status, current, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health sca le, d M3 = School health self efficacy questionnaire,

PAGE 84

74 e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***co rrelation is significant at <.001 level. Figure 8 . Diagram of psychological mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors. Note . a X = Subjective Social Status, future, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 = Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 9 . Diagram of psychological mediators on relationship between access to health resources and maternal lifestyle beh aviors. Note . a X = Access to health resources, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d M3 = School health self efficacy questionnaire, e M4 =

PAGE 85

75 Coping inventory for stressful situations, task oriented subscale, f Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 10 . Diagram of maternal depression and learned helplessness media tors on relationship between economic hardship and maternal lifestyle behaviors. Note . a X = Economic Hardship Scale, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 11 . Diagram of maternal depression and learned helplessness mediators on relationship between educational achievemen t and maternal lifestyle behaviors. Note . a X = Years of education, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlat ion is significant at .01 level, ***correlation is significant at <.001 level.

PAGE 86

76 Figure 12 . Diagram of maternal depression and learned helplessness mediators on relationship between annual personal income and maternal lifestyle behaviors. Note . a X = Annua l personal income, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation i s significant at <.001 level. Figure 13 . Diagram of maternal depression and learned helplessness mediators on relationship between annual family income and maternal lifestyle behaviors. Note . a X = Annual family income, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.0 01 level.

PAGE 87

77 Figure 14 . Diagram of maternal depression and learned helplessness mediators on relationship between social support and maternal lifestyle behaviors. Note . a X = Multidimensional Scale of Perceived Social Support, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 15 . Diagra m of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, current, and maternal lifestyle behaviors. Note . a X = Subjective Social Status, current, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.0 01 level.

PAGE 88

78 Figure 16 . Diagram of maternal depression and learned helplessness mediators on relationship between Subjective Social Status, future, and maternal lifestyle behaviors. Note . a X = Subjective Social Status, future, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 17 . Diag ram of maternal depression and learned helplessness mediators on relationship between access to health resources and maternal lifestyle behaviors. Note . a X = Access to health resources, b M1 = Multidimensional health locus of control, internal subscale, c M 2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.

PAGE 89

79 Figure 18 . Diagram of psychological mediators on relationship between inability to make ends meet subset of EHS and maternal lifestyle behaviors. Note . a X = Inability to make ends meet average, b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Figure 19 . Diagram of psychological mediators on relation ship between adjustments and cutbacks subset of EHS and maternal lifestyle behaviors. Note . a X = Adjustments and cutbacks count , b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyl e Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level.

PAGE 90

80 Figure 20 . Diagram of psychological mediators on relationship between not enough money for necessities of EHS an d maternal lifestyle behaviors. Note . a X = Not enough money for necessities average , b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is significant at <.001 level. Fig ure 21 . Diagram of psychological mediators on relationship between financial strain subset of EHS and maternal lifestyle behaviors. Note . a X = Financial strain average , b M1 = Multidimensional health locus of control, internal subscale, c M2 = Value on health scale, d Y = Health Promoting Lifestyle Profile *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlation is s ignificant at <.001 level.

PAGE 91

81 Figure 22. prediction of maternal lifestyle behaviors. *Correlation is significant at .05 level, **correlation is significant at .01 level, ***corr elation is significant at <.001 level. Figure 23. maternal lifestyle behaviors through mediator maternal depression. *Correlation is significant at .05 level, **correlation is si gnificant at .01 level, ***correlation is significant at <.001 level.

PAGE 92

82 Figure 24. maternal lifestyle behaviors through mediators of binary variables for years of education and r ace/ethnicity. *Correlation is significant at .05 level, **correlation is significant at .01 level, ***correlatio n is significant at <.001 le ve l.