Citation
Health at the intersection of work and family for American mothers

Material Information

Title:
Health at the intersection of work and family for American mothers
Creator:
Finnigan-Fox, Grace
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 Sociology
Degree Disciplines:
Sociology
Committee Chair:
Lippert, Adam M.
Committee Members:
Cooney, Teresa M.
Alexander, Kari

Notes

Abstract:
The combination of work and motherhood is increasingly common in the United States, with over half of mothers employed in some capacity. However, research to date has focused pri-marily on broad aspects of both work (eg, employed vs. unemployed), and family life (married vs. unmarried); the health implications of detailed combinations of employment and family cir-cumstances have received less attention, both in respect to mental as well as behavioral and physiological well-being. Understanding the relationships among domains of psychological dis-tress, behavior, and physiological functioning earlier in the life course may benefit disease pre-vention initiatives. This study aims to address these gaps by using data from the National Longi-tudinal Study of Adolescent to Adult Health (Add Health) to explore 1) how detailed aspects of work-family circumstances relate to perceived stress and depression among young working mothers, 2) whether detailed work-family characteristics that bear upon stress and depression are similarly related to a vector of coping behaviors, and 3) whether any aspects of work-family cir-cumstances found to be correlated with stress, depression and coping behaviors are also corre-lated with biomarkers gauging cardiometabolic health. Results from multivariable regression models show that certain work characteristics (such as repetitive work) are significantly associ-ated with multiple dimensions of health, while others (such as decision-making latitude) are as-sociated with just one dimension (eg, self-reported stress) but not another. Taken together, the results from this study illustrate the nuanced ways in which work and family life combine to in-fluence physiological and mental well-being, as well as the health behaviors that often compound these health outcomes.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Grace Finnigan-Fox. 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 is only available as the following downloads:


Full Text

PAGE 1

HEALTH AT THE INTERSECTION OF WORK AND FAMILY FOR AMERICAN MOTHERS by GRACE FINNIGAN FOX B.A., Berklee College of Music, 2010 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 Sociology 2018

PAGE 2

This thesis for the Master of Arts degree by Grace Finnigan Fox has been approved for the Sociology Program by Adam M. Lippert, Chair Teresa M. Cooney Kari Alexander Date: December 15, 2018

PAGE 3

ii FinniganFox, Grace (MA, Sociology) Health and the Intersection of Work and Family for American Mothers Thesis directed by Assistant Professor Adam M. Lippert ABSTRACT The combination of work and motherhood is increasingly common in the United States, with over half of mothers employed in some capacity. However, research to date has focused primarily on broad aspects of both work (eg, employed vs. unemployed), and family life (married vs. unmarried); the health implications of detailed combinations of employment and family circumstances have received less attention, both in respect to mental as well as behavioral and physiological well being. Understanding the relationshi ps among domains of psychological distress, behavior, and physiological functioning earlier in the life course may benefit disease prevention initiatives. This study aims to address these gaps by using data from the National Longitudinal Study of Adolescen t to Adult Health (Add Health) to explore 1) how detailed aspects of work family circumstances relate to perceived stress and depression among young working mothers, 2) whether detailed work family characteristics that bear upon stress and depression are s imilarly related to a vector of coping behaviors, and 3) whether any aspects of work family circumstances found to be correlated with stress, depression and coping behaviors are also correlated with biomarkers gauging cardiometabolic health. Results from m ultivariable regression models show that certain work characteristics (such as repetitive work) are significantly associated with multiple dimensions of health, while others (such as decision making latitude) are associated with just one dimension (eg, sel f reported stress) but not another. Taken together, the

PAGE 4

iii results from this study illustrate the nuanced ways in which work and family life combine to influence physiological and mental well being, as well as the health behaviors that often compound these he alth outcomes. The form and content of this abstract are approved. I recommend its publication. Approved: Adam M. Lippert

PAGE 5

iv TABLE OF CONTENTS CHAPTER I. INTRODUCTION.............................................................................................1 II. BACKGROUND...............................................................................................4 Mothers in the Workforce..................................................................................4 Role Strain and Enhancement for Working Mothers........................................6 Stress, Coping, and Physical Health of Working Mothers................................9 III. METHODS.......................................................................................................14 Measures...........................................................................................................15 Psychological distress.................................................................................15 Health Behaviors........................................................................................16 Cardiometabolic Biomarkers......................................................................17 Covariates...................................................................................................18 Analyses............................................................................................................20 IV. RESULTS.........................................................................................................21 Work Family Characteristics and Stress and Depress ion Scores.....................21 Work Family Characteristics and Engagement in Health risk behaviors ........22 Work Family Characteristics and Biomarkers of Metabolic Dysfunction.......24 V. DISCUSSION...................................................................................................25

PAGE 6

v VI. REFERENCES.................................................................................................28 LIST OF TABLES TABLE 1. Descriptive Statistics..............................................................................................33 2. Work , Family and Psychological Distress.............................................................34 3. Work, Family and Health risk behaviors ...............................................................35 4. Work, Family, and Cardiometabolic Health..........................................................36

PAGE 7

1 CHAPTER I INTRODUCTION Cardiometabolic diseases are among the leading causes of death in the U.S. (Kochanek, Murphy, Xu, & Arias, 2017). These diseases are also highly preventable, with engagement in health risk behaviors such as smoking, drinking, and poor diet contributing to their risk (Benjamin et al., 2017; S. L. Jackson, King, Zhao, & Cogswell, 2016; Warren, Alberg, Kraft, & Cummings, 2014). The CDC has estimated that at least 200,000 deaths due to heart or cardiovascular disease among persons under the age of 75 could be prevented annually simply by altering health behaviors (National Center for Chronic Disease Prevention and Health Promotion, 2013). While there have been some encouraging declines in health risk behaviors, such as decreases in tobacco use (CDC, 2016), the threat to population health posed by these behaviors is still co nsiderable: deaths attributable to smoking remain high (Warren et al., 2014), and the prevalence of diet related diseases in the U.S., namely diabetes, has increased since 1988 (Menke, Casagrande, Geiss, & Cowie, 2015). While health risk behaviors heighten one’s risk for poor cardiometabolic health, they may also provide temporary relief from life stress. Although the stress health behavior relationship is complex, with some literature suggesting that health behaviors drive stress levels (MacFarlane & Montg omery, 2010), the most commonly supported hypothesis is that health behaviors serve as a coping mechanism in response to stressful experiences (Park & Iacocca, 2014). Smoking, alcohol consumption, and diets high in sugar and fat have been shown to dampen e xperiences of psychological distress (Dallman et al., 2003). Further, those living in chronically stressful situations are particularly prone to coping responses featuring health risk behaviors (J. S. Jackson, Knight, & Rafferty, 2010).

PAGE 8

2 One critical sourc e of chronic stress occurs at the intersection of work and family circumstance, particularly for working mothers. Women’s life courses are more complex than men’s, and more often feature stressful work family circumstances like single parenting while worki ng. (Livingston, 2013). Although combining work and family has generally been found to be beneficial for mothers (Frech and Damaske, 2012), the competing demands of both can be emotionally and physically taxing. On the one hand, continued engagement in wo rk outside the home is beneficial for mothers’ psychological and physiological health as well as their economic security (Damaske, 2011; Frech & Damaske, 2012). However, problems may arise when work or family roles demand more resources than are available: mothers who must manage work family conflict and family work conflict are more likely to experience psychological distress (Hill, 2005) and engage in health risk behaviors , such as smoking or alcohol consumption (Frone, Barnes, & Farrell, 1994). In sum , sociological research has shown that combining work and family is generally beneficial for women, but that the demands placed upon working mothers by competing work family roles engender conflicts that may be stressful to manage. Further, engagement in health risk coping behaviors may offer temporary respite from these work family stressors; however, they also increase one’s risk of developing cardiometabolic disease. Despite these important findings, gaps regarding the relationship between work, family a nd health for mothers remain. First, extant research has focused primarily on broad aspects of both work (employed women vs. women who do not work), and family life (mothers vs. childfree women). The question of how detailed aspects of mothers’ employment —decision making authority, hours worked, physical exertion required by one’s work —are associated with health has received less attention. Second,

PAGE 9

3 little work has been done to link these detailed aspects of mothers’ work to both mental, behavioral, and car diometabolic health. What research exists has been limited to one or two outcomes, such as physical and mental health (Frech and Damaske, 2012) or health behaviors (Frone, Barnes and Farrell, 1994). Third, prior work has focused much attention on later lif e, although early adulthood is a life course stage where work and family circumstances begin to shape health behaviors. To address these gaps, this study draws from sociological theories on work, family, and the stress process model. Using these theories, this study investigates the following aims: 1) whether there are associations between detailed aspects of work circumstances and mental, behavioral, and cardiometabolic health in a sample of young adult working mothers, 2) whether these detailed occupational characteristics associate differently with separate facets of mental health (self reported stress and depression), and 3) whether adjustments f or health ri sk coping behaviors and occupational characteristics alter the association between mental strain and cardiometabolic health among working mothe rs. Data from wave IV of the National Longitudinal Study of Adolescent to Adult Health (Add Health) are utilized for this purpose. Findings from this study highlight how aspects of work that ease or exacerbate work family conflict are associated with work ing mothers’ health in young adulthood.

PAGE 10

4 CHAPTER II BACKGROUND Mothers in the Workforce The typical mix of family and employment circumstances for U.S. women began to change in the early 20th century. Multiple social forces, from the women’s suffrage movement to the shortage of male laborers during the first World War, combined to unlock educational and employment opportunities that were previously blocked (DOL, 2012). Since then, the proporti on of the workforce constituted by women has increased from 21% in 1920 to 47% in 2010 (DOL, 2012). While the changes in women ’s work force participation have been profound, shifts in the gendered division of domestic labor have been more limited. Motherhood is still considered a normative experience for many women, and mothers still shoulder the bulk of household labor and manage core aspects of childrearing, such as feeding and dressing (Bianchi, Sayer, Milkie, & Robinson, 2012; Musick, Meier, & Flood, 2016). Further, the “intensive mothering” model of parenting demands mothers focus all of their time and energy on raising a high quality child (Hays, 1996). These cultural prescriptions leave mothers’ parenting decisions under close scrutiny; everything fr om birthing to breastfeeding decisions are highly politicized and publicly critiqued (Feeley & Thomson, 2016; Wall, 2001). The decision to work in particular presents a dilemma to mothers and exposes them to social criticism. Working mothers frequently suf fer from guilt or shame due to spending time away from their children, and often build complex narratives accounting for and justifying this decision (Damaske, 2013; Sutherland, 2010). Further, social

PAGE 11

5 narratives surrounding challenges facing children and a dolescents today, like the pediatric obesity epidemic, have frequently hinged on mothers’ employment as a cause (Martin, Lippert, Chandler, & Lemmon, 2018). Despite the intense household demands and social sanctions placed upon them, the majority of mother s do participate in the labor force; over 70% of mothers with children under the age of 18 in the house were simultaneously performing paid labor in 2013 (DOL, 2014). This holds true for mothers of young children as well: since 1968, the proportion of moth ers with children under the age of one who participate in the labor force has risen from 21% to roughly 50% from 1986 onward (Han, Ruhm, Waldfogel, & Washbrook, 2008). Similarly, job continuity is strong among women who worked full time prior to pregnancy, with the majority returning to their pre pregnancy employers after maternity leave (Klerman & Leibowitz, 1999). Thus, combining work with motherhood is the modal experience for mothers in the U.S. Despite the fact that combining work and family has becom e commonplace, mothers continue to be systematically disadvantaged in the workplace when compared to their childfree and male counterparts. Although mothers are now breadwinners for 40% of all U.S. households (Wang, Parker, & Taylor, 2013), and continue to manage the bulk of domestic labor and childcare (Bianchi et al., 2012; McDonald, 2000), their earnings pale in comparison to their male counterparts, with mothers earning 74.7 cents to every working fathers’ dollar (ACS, 2012). Similarly, the family wage gap, where in childfree women earn more than mothers, persists, though evidence suggests this is changing among those in highly skilled professions (Buchmann & McDaniel, 2016). These inequalities engender economic disparities between men and women, mothers and nonmothers, and among working mothers in well and poorly compensated posi-

PAGE 12

6 tions. In addition to economic disparities, there is considerable variation in employment circumstances among working mothers. As a result of increasing polarization in the wor kforce between skilled and unskilled employment, working mothers encounter a variety of employment circumstances, differentiated by class, that may exacerbate tensions between work and family roles for some and lead to health inequalities (Kalleberg, 2011) . F i nally , working mother’s family circumstances may s ignificantly shape their career trajectories . As theorized by Moen and Chermack (Moen and Chermack, 2005), expectations around what constitutes a typical career path still cleave to a traditionally maleas breadwinner model, in which additional family centered work is not accounted for. The intersection of women’s occupational and family centered work are not taken in to account; as such, working mothers must contend not only with ques tions of economic circumstances, but with issues of family structure and timing as well when considering employment opportunities (Moen and Chermack, 2005). Role Strain and Enhancement for Working Mothers Prior sociological work has focused primarily on b road aspects of work family circumstances and health outcomes, such as comparisons between working mothers and working fathers, working and nonworking mothers, and labor force participation by women with and without children (Frech & Damaske, 2012; Musick et al., 2016; Yua & Kuo, 2017). However, less work has been devoted to detailed aspects of work family combinations and their implications for health solely for working mothers. As previously discussed, this population faces significant demands from both work and family, suggesting that the work family experiences of working mothers are particularly unique and complex. Further, these competing demands —along with the social sanctions leveled at mothers who manage both—place working mothers at high risk for

PAGE 13

7 role strain . The theory of role strain proposes that all persons perform multiple, frequently competing roles, and that the material, social, and emotional resources necessary to satisfy the demands of each role are limited, making it impossible to fully c omply with all of one’s roles (Goode, 1960). The struggle to meet the requirements of all of one’s roles results in psychological distress (Goode, 1960). The claims by work and family placed upon mothers’ resources, especially when combined with insufficie nt economic and social supports, certainly do make working mothers more vulnerable to strain. Research on work family conflict has shown as much, with working mothers at higher risk for work family conflict, and concomitant psychological distress, than wor king fathers (Hill, 2005). However, despite this increased vulnerability to work family conflict, a larger body of research has found that combining work and family roles is beneficial for mothers. Working mothers enjoy better long term physical and mental health than nonworking mothers (Frech and Damaske, 2012; Damaske, 2011). Further, the workplace provides relief for some mothers from household stressors (Damaske, Smyth, & Zawadzki, 2014; Hochschild, 1997). These findings are consistent with the theory of role enhancement, which posits that performing multiple roles results in desirable social and economic rewards, such as increased income, perceived mastery, or self esteem (Marks, 1977; Sieber, 1974). The benefits attained by managing multiple roles outweigh any strain, resulting in net psychological benefit (Sieber, 1974). In sum, while working mothers face significant challenges as a direct result of combining work and motherhood, the majority of U.S. mothers continue to participate in the labor force , and the majority of research has found that employment confers psychological and physical ben-

PAGE 14

8 efits to mothers. However, most of this research has focused on broad aspects of work; it is unclear whether and how the health benefits of combining work and fa mily are influenced more specifically by detailed aspects of one’s work. Some insight on how detailed aspects of working mothers’ employment influence psychological, physical, and behavioral health may be gleaned from Karasek’s demandcontrol model (Karasek, 1979). Under this model, the “ demands ” of work are multidimensional, and constitute a matrix of interdependent experiences that outline the expectations placed upon workers (Karasek, 1979). Key characteristics include decisionmaking latitude, job con trol, and monotony of work. Combinations of these characteristics influence employee’s psychological well being, with more unbalanced combinations (eg, demanding work with little control) conveying greater risk of poor mental health. However, while high de mand, low control jobs —or “high strain” jobs —may be the most deleterious to health, low demand, low control (or “ passive ”) jobs may also be detrimental to health. Karasek posited that passive jobs led to a decline in problem solving ability and cognitive activity, and resulted in a “learned helplessness” (Karasek, 1979). Further, the health implications of high strain and passive jobs do not end with mental health — employees in high strain or passive jobs are also at a greater risk of engaging in risk behav iors such as heavy alcohol use (Wiesner, Windle, & Freeman, 2005). An extended version of the demand control model, which includes family based social support, further illuminates the complexity of combining work and family. The work family strain model en visions health as being predicated upon three dimensions: demands from both work and family, control over workplace circumstances, and formal or informal social support (Berkman, 2015). Workers with the highest demands from work and family, lowest control over

PAGE 15

9 the circumstances of their employment, and least informal or formal social support are posited to have the poorest health (Berkman, 2015). Research has found the employment characteristics detailed in the demand control and work family strain models are linked not only to psychological strain, but to higher rates of morbidity and mortality as well. A salient example is the Whitehall II study, which followed British civil servants for three years and tracked both employment status and health. A crucial finding was that civil servants with lower status jobs (jobs offering lower control and prestige) had worse self reported health and more symptoms of cardiometabolic disease (Marmot et al., 1991). Similarly, a study specifically focused on disability and female workers found that jobs that were intellectually stimulating were associated with lower likelihood of physical disability later in life, compared to jobs that were monotonous or physically taxing (Palumbo et al., 2017). While some research has exami ned the health of working mothers through the lens of either the demand control model or work family strain model, the focus has primarily been on a single explanatory characteristic, such as job control or outside social support, with mortality and long t erm health as an outcome (Robinson, Magee, & Caputi, 2016; Sabbath, Mejia Guevara, Noelke, & Berkman, 2015). As a result, the relationships among detailed work characteristics, physiological, mental, and behavioral health among younger working mothers has received inadequate attention. Stress, Coping, and Physical Health of Working Mothers Considering one major finding of the Whitehall II study was that employees in lower status jobs also engaged in more health risk behaviors (Marmot et al, 1991), the na rrow focus of prior research on either physiological, mental, or behavioral health outcomes ignores how all

PAGE 16

10 three may be similarly associated with workplace exposures. Whitehall II illuminated the association between low status, low control jobs and engagement in health risk behaviors (Marmot et al, 1991). Similarly, lack of schedule control has been tied to engagement in health risk behaviors . Research examining the relationship between employee health and frequent overtime work found that frequent overtim e was associated with both poor self rated health and poor diet (Taris et al., 2011). One hypothesis for this relationship between less favorable workplace circumstances, such as poor schedule control or low status, and higher engagement in health risk behaviors is that adverse workplace conditions rob employees of the positive coping mechanisms used to engage in healthier stressrelieving a c tivities , such as physical activity or meal preparation, leaving employees with no choice but to turn to easily acc essible but low quality alternatives (Romelsjo et al., 1992; Taris et al., 2011). Support for this hypothesis can be found in broader research on the stress process and coping mechanisms. Pearlin’s stress process model suggests that there are three core domains of the stress experience: the cause of stress, the mediators and moderators of stress, and the manifestations of stress (Pearlin, Lieberman, Menaghan, & Mullan, 1981). Pearlin hypothesized that sources of stress arose from the context of one’s life e xperiences, and could compound one another, resulting in disparities: for one person, a stressful life event might not trigger a stressful response, but for another, a stressful life event might serve to exacerbate the effects of other stressors, prompting a much stronger negative response (Pearlin et al., 1981). Pearlin posited that persons in such stressful circumstances had two primary means of mediating said stress: social supports and coping (Pearlin et al, 1981). Coping is a behavior born of the indi vidual’s desire to modify some aspect of the stressful situation so as to reduce its perceived threat or impact (Pearlin et al, 1981). A growing body of work has pointed to health risk

PAGE 17

11 behaviors as one such form of coping. Health risk behaviors have been c onceptualized as avoidant coping mechanisms, attenuating the psychological experience of stress without actually addressing its source (Mezuk et al., 2017). Persons in chronically stressful situations often engage in health risk behaviors , such as smoking, excessive alcohol intake, and consumption of fat and sugar heavy foods (Jackson et al, 2010). Engagement in these behaviors has been shown to dampen experiences of stress and depression by acting through the hypothalamic pituitary adrenal (HPA) axis to ei ther moderate the release of stress hormones or increase the release of dopamine and beta endorphin levels (Dallman et al., 2003; Kirschbaum, Wust, & Strasburger, 1992; Peele & Brodsky, 2000). Further, longitudinal studies have found a strong, positive rel ationship between perceived stress, health risk behaviors and cardiac risk profile (Rod, Gronbaek, Schnohr, Prescott, & Kristensen, 2009).Thus, while these health risk behaviors may provide perceived relief from stress in the short term, the longterm effects of chronic stress, and frequently attendant coping behaviors, has significant deleterious effects to ones cardiometabolic health. Evidence of engagement in health risk behaviors as a coping mechanism for stress can be found within the context of work family strain as well. Work family conflict is positively associated with health risk behaviors such as smoking (Nelson, Li, Sorensen, & Berkman, 2012), poor diet (Allen & Armstrong, 2006), and heavy alcohol use (Frone, Russell, & Barnes, 1996). Strain between work and family has also been linked to substance use, such as cigarette smoking and heavy drinking, among employed mothers (Frone et al., 1994). However, less is known about how an array of work and family characteristics that directly influence psychological distress similarly bear upon health behaviors that are also tied to both poorer mental health and cardiovascular risk profile.

PAGE 18

12 In sum, the literature on work family confli ct and strain identifies several key findings: 1) while employment in general confers socioeconomic, psychological, and physical health benefits to mothers, it is not an unconditional benefit, with the rewards reaped from combining work and family dependen t on detailed aspects of each domain; 2) the competing theories of role strain and role enhancement support this tenuous balance between the benefit of accruing multiple roles, and the potential for distress that occurs if these roles demand more than one can give; 3) that persons in chronically stressful environments often engage in coping mechanisms —such as health risk behaviors —that may narrow certain health disparities (e.g., mental health inequalities) and widen others (e.g., cardiometabolic health). Despite these insights on the relationships between work, family and health, crucial gaps in knowledge remain. First, the bulk of current research on the relationship around work and family characteristics and health has primarily compared and contrasted m others and fathers, working mothers and nonworking mothers, and working women with and without children. Given that combining work and motherhood is the modal experience for U.S. mothers, a more nuanced exploration of the health disparities that track wit h detailed aspects of work and health among mothers is important. Second, studies of the relationships between work, family, and health often focus on detailed aspects of family life, such as marital status or social support, with less attention paid to de tailed aspects of work. T aking into consideration the importance of family structure and timing on employment choices or options for working mothers (Moen and Chermack, 2005), controlling for these family variables while exploring the relationship between detailed aspects of work and health outcomes may provide greater insight in to how these work characteristics influences health for working mothers, regardless of their family circumstances. Third, most studies of work, family and health have a

PAGE 19

13 single mental health outcome, most frequently depression. Considering that both stress and depression, despite being separate constructs, are associated with health risk behaviors and cardiometabolic disorder (Rod et al., 2009; Suls & Bunde, 2005) the focus on a single measure of menta l strain as an outcome is a limitation. Finally, most research focuses on long term effects of occupational characteristics and family on the health of working mothers. Thus, less is known about this relationship with health at an earlier stage in life, prior to the em ergence of clinically identifiable chronic disease. This study aims to fill these gaps by examining 1) how family circumstances and workplace decisionmaking authority, job monotony, shift work, and other detailed aspects of employment are associated with young working mothers’ psychological well being, 2) whether these same aspects of work and family are associated with health risk behaviors including tobacco use, alcohol consumption, and diet, and 3) whether biomarkers gauging cardiometabolic disorders high bloodpressure, HbA1c levels, and body mass index are significantly associated with specific aspects of work or family with and without adjustments for self reported measures of stress and depression.

PAGE 20

14 CHAPTER III METHODS Data are drawn from the National Longitudinal Study of Adolescent to Adult Health (Add Health). Add Health was initiated in 1994 to investigate the health and risk behaviors among young people as they transition from adolescence to adulthood, with a primar y focus on health outcomes (both self reported health and physical specimens), and health behaviors, family structure and working life. Add Health is the largest and most comprehensive longitudinal study of adolescents in the United States. Participants surveyed during Wave IV of Add Health were between the ages of 24 and 32, ages where many young adults begin to take on careers and begin families, with a total sample size of N=15,701. The analytic sample for this study was restricted to women (n=8,352) who indicated they had at least one child at home (n=4,731) and worked ten or more hours a week (n=3,333). Due to the difficulty of imputing on biological variables, listwise deletion was used for any cases of missing data for all biomarker data. I used Stata ’s ICE program (StataCorp LP, College Station, Texas) to multiply impute any other missing data for key independent and dependent variables (Royston, 2004). 43% of all cases were missing data; as such, 43 data sets were imputed (Bodner, 2008). The imp uted datasets of complete data were then combined to adjust for within and between variance in the imputed samples (Acock, 2005). The fraction of missing information (FMI) aligned with recommended values for post imputed data in all but one case (alcohol consumption). The final sample size utilized in this analysis was N=3,227. Survey weights were applied post imputation to adjust for the survey design.

PAGE 21

15 Measures Psychological distress Self reported mental health is comprised of two subdomains: perceiv ed stress and perceived depression. Perceived depression was measured using the 10Item Center for Epidemiological Studies Depression Scale (CES D). During the Wave IV in home survey, individuals were asked to indicate how often they had experienced each of the following in the past seven days: a) were bothered by things that usually don’t bother them, b) could not shake off the blues, even with help from family and friends, c) felt they were as good as other people (reverse coded), d) had trouble keeping t heir mind on what they were doing, e) felt depressed, d) felt too tired to do things, e) felt happy (reverse coded), f) enjoyed life (reverse coded), g) felt sad, and h) felt that people disliked them. Responses ranged from 0 (never or rarely) to 3 (most or all of the time), with higher overall scores indicating higher levels of de pression. Respondents’ scores are the sum of their answers to all ten questions. The resulting measure has a range of 0 to 30. The results were slightly skewed; however, sensitivity tests were run, and log transformation of the variable did not improve the measure. Therefore, the nontransformed scale was used in analyses. Perceived stress was measured using Cohen’s 4Item Perceived Stress Scale (PSS4). Respondents were asked to in dicate how often in the past 30 days they had: a) felt they were unable to control the important things in their life, b) confident in their ability to handle their personal problems (reverse coded), c) felt things were going their way (reverse coded), and d) felt that difficulties were piling up so high that they could not overcome them. Responses for each item ranged from 0 (never) to 4 (very often), with higher scores indicating higher levels of perceived

PAGE 22

16 stress. This variable was operationalized as a c ontinuous variable, with total possible scores ranging from 0 to 16. This variable was normally distributed; thus, no sensitivity tests were run. Health Behaviors Health risk behaviors included smoking, binge drinking, fast food consumption, and consumpti on of sugary drinks. Smoking was assessed using the question, “during the past 30 days, on how many days did you smoke cigarettes?” Answers were coded 1 if the respondent indicated they smoked on one or more days in the past 30 days, and 0 otherwise. Similarly, respondents were asked about their alcohol consumption in the past 30 days, specifically how many drinks they imbibed each time they consumed an alcoholic beverage. Binge drinking was operationalized as a dichotomous variable, with participants who consumed 4 or more drinks at a sitting coded as 1—indicating they’d had at least one binge drinking episode in the past 30 days —and 0 otherwise. These cut offs were determined based off of the National Institute of Health’s (NIH) definition of binge drinki ng for women. Prior to dichotomization, means number of drinks consumed at each sitting in the past 30 days was 2.1, with a minimum of 0 drinks and a maximum of 18 drinks. Questions assessing fast food and sugary drink consumption in the past seven days we re used to create an ordinal index of poor diet. Respondents who reported consuming both fewer than three sugary drinks and fewer than four fast food meals were coded as 0; those who reported consuming three or more sugary drinks or four or more fast food meals were coded 1; and those who reported three or more sugary drinks and four or more fast food meals were coded 2. These cut offs were determined based off of prior research showing that three or more sugary drinks or four or more fast food meals per we ek were correlated with worse health (Miedema et al., 2015).

PAGE 23

17 Cardiometabolic Biomarkers Biomarkers of cardiometabolic health were measured using a respondent’s blood pressure class, their body mass index, and diabetic status. All three measures were created based off of measurements taken by trained Add Health study personnel, after which ordinal scales were constructed based off of approved cut points. Blood pressure class was measured ordinally as normal, pre hypertensive, and hypertensive. These classifi cations were defined using guidelines from the Seventh Report of the Joint National Committee on Prevention Detection, Evaluation, and Treatment of High Blood Pressure (Chobanian et al., 2003). Respondents placed in the normal blood pressure class if their systolic blood pressure (SBP) was less than 120 mmHg and diastolic blood pressure (DBP) was less than 80mmHg. Those classified as pre hypertensive had an SBP between 120139 mmHg and a DBP between 8089 mmHg, and those classified as hypertensive had an SB P of greater than 140 mmHg and a DBP of greater than 90 mmHg. BMI was a constructed variable using the participant’s height and weight. Recommended cut points from the CDC (CDC, 2014) were used to create an ordinal variable with three levels: underweight or normal weight (BMI or 24.9 kg/m2 or less), overweight (BMI of 2529.9 kg/m2 ) or obese (greater than 30 kg/m2). Diabetic status was also a constructed variable. A respondent was classified as diabetic, pre diabetic, or diabetic based off of HbA1c readings. An HbA1c of less than 5.7% was categorized as normal; 5.76.4% was categorized as pre diabetic; and greater than 6.4% was categorized as diabetic. These categories were constructed using previously identified cut points (American Diabetes, 2015).

PAGE 24

18 Covariates Independent variables utilized in this study fall in to two domains: labor experiences and family structure. Labor experiences include physical exertion required by the work performed, full time employment, whether the participant worked conve ntional hours, decision making freedom, occupational classification, and how repetitive the work was. Physical exertion required by the participants’ job was measured using responses to the following: In your current job, do you spend most of your time 1) standing, doing hard physical work, for example doing construction work; 2) standing doing moderate physical work, for example, nursing or being a mechanic; 3) standing, doing light physical work, for example, standing at a counter, teaching or working at a conveyer belt; 4) seated, for example, using a computer or driving. Participants were classified as having physically demanding jobs if they selected either of the first two responses (=2); jobs that required little physical exertion if they selected the third response (=1); and jobs that required no physical exertion if they selected the fourth response (=0). Work hours were measured as part time (less than 30 hours a week), full time (30 40 hours a week) and over time (more than 40 hours a week). These classifications were based on guidelines from the Affordable Care Act (ACA) and the Fair Labor Standards Act (FLSA). Type of shift work engaged in by the participant was measured using the question: “which of these categories best describes the hours you w ork at your job? 1) regular day shift 2) regular evening shift 3) regular night shift 4) shift rotates —that is, it changes periodically from day to evening or night 5) split shift —that is, it consists of two distinct periods each day 6) irregular schedule or hours 7) other”. Participants were categorized as working a regular day shift (=0); evening shift (=1) or irregular hours (=2). Decisionmaking freedom at work was measured ordinally using the question: “Overall, how often do you have the freedom to ma ke important decisions about what you do at work and how you do

PAGE 25

19 it? 0) none or almost none of the time, 1) some of the time, 2) most of the time, 3) all or almost all of the time”. Due to a small number of respondents for “none of the time”, “some of the t ime” and “none of the time” were combined. Occupational classification was measured using the question, “now I'd like to record a description of your (current/most recent) job. When you see the list of categories, please tell me which best describes what you (do/did) at your (current/most recent) job”. Using guidelines from the Bureau of Labor Statistics (BLS), participants were categorized as either falling in to the management, business, science, and arts category (=0), or the service/sales clerical/natu ral resources/construction and maintenance/production, transportation and material moving category (=1). Although the BLS aggregates occupations from the second group in to several separate categories, they are grouped together for the purposes of this stu dy due to the fact that these occupations are less frequently held by women, and would prove to be less statistically powerful on their own. Finally, repetition of work was measured ordinally using the question, “how much of the time do you do the same thi ngs repeatedly, that is over and over? 0) none or most of the time, 1) some of the time, 2) most of the time, 3) all or almost all of the time”, with higher values indicating more repetitive work. Similar to decision making latitude, due to a small number of respondents for “none of the time”, “some of the time” and “none of the time” were combined for repetitiveness of work as well. Family structure included partner status, number of children in the home, and personal income. Partner status was measured using a question that asked about members of the respondents ’ household. Respondents were then classified as married (=2), cohabiting (=1), or single (=0). Similarly, number of children in the home was measured as a count of persons of any age indicated to be sons/daughters by the respondent when asked to detail the members of their household. Personal income was based off a question asking how much the respondent made per

PAGE 26

20 year; the question was then sorted into quartiles ($10,000 or less; more t han $10,000 or $20,000 or less; more than $20,000 or $30,000 or less; more than $30,000 or $40,000 or less; more than $40,000). Several control variables related to cardiometabolic health and self reported stress/depression were also included. Race was m easured categorically as “nonwhite” (=1), and white (=0). Of note, a question regarding ethnicity was not present; therefore, it was not possible to tell whether those who responded as white were nonHispanic or Hispanic white. Income in Add Health is measured in brackets; as such, the median value of each bracket was used to generate a continuous variable for respondent income. Analyses Simple descriptive tests were run post imputation to determine the characteristics of the sample. Multivariate analyses were then run for the three outcome domains. Each outcome domain included multiple models: the domain of mental health included models w ith just family variables, just work variables, and both; health behavior and biomarker domains included models with and without mental health variables (stress and depression). Ordinary least squares (OLS) regression was used in cases of linear outcomes, while logistic and ordered logistic regression were used to measure, respectively, binary and ordinal outcomes.

PAGE 27

21 CHAPTER IV RESULTS Table 1 provides descriptive statistics for the sample. Briefly, 54.2% of mothers were married, 17.96% were cohabiters, and 27.84% were single, and the mean number of children in a woman’s care was 1.84. 16.3% worked between 10 and 30 hours a week, 64.77% worked between 30 and 40 hours a week, and 18.93% worked more than 40 hours a week. Mean age was approximately 29 ye ars, with a mean household income of $56,774 a year. 75.03% of the sample was Hispanic or nonHispanic white, and 65.01% had attended college. Work Family Characteristics and Stress and Depression Scores The first set of models examines the relationship between several characteristics of employment circumstances, family life, and self reported stress and depression. Three models were run: one with just employment variables, one with just family variables, and one with both. The results of these analyses m ay be seen in Table 2. Model 1, which included only family variables and controls, is not shown for the sake for parsimony. In model 2, controlling for all other employment and control variables, those with a high degree of decisionmaking latitude had a s tress score that was .62 points lower than those with only some or no decision making latitude. The difference between those with high decision making latitude and low decision making latitude was attenuated upon the inclusion of family characteristics, to a difference of .61 points, though the coefficient for decision making latitude remained significant. Decision making latitude had no significant relationship with depression scores; however, repetitiveness of work did. Those whose work was repetitive all of the time compared to none or some of the time had a depression score that was .58 points higher, and those whose work was repetitive most of the time had a

PAGE 28

22 score that was .35 points higher. Upon inclusion of family variables in the model, the relations hip between repetitiveness of work and depression score shrank to a difference of .53 points between those whose work was repetitive all of the time vs. none of the time, and .32 points for those whose work was repetitive most of the time vs. none of the t ime. Repetitiveness of work was not, however, related to stress scores. Finally, full time workers (30 40 hours a week) enjoyed a stress score that was .45 points lower than those who worked part time; however, this difference was attenuated by the additi on of family variables. Hours worked were not significantly associated with depression scores. Turning to the family variables in Table 2, partner status, specifically marital status, and number of children were significantly and positively correlated w ith increased scores in both the family only and the family work model for both stress and depression. When adjusting for income, race, education and other family and work characteristics, married mothers enjoyed a stress score that was .70 points lower th an their single counterparts, a nd a depression score that was .65 points lower. Controlling for the same set of variables above, model 2b indicates that, for each additional child, stress scores rose by .24 points; number of children, however, was not s ignificantly associated with depression scores. While marital status was significantly associated with both psychological outcomes, working mother’s personal earnings were not in either of the models. Work Family Characteristics and Engagement in Healt h Risk B ehaviors Results shown in Table 3 assess the relationships between work and family characteristics and a series of health risk behaviors , with and without adjustment for stress and depression. Decision making latitude was not significantly associated with engagement in any of the health behaviors, but mothers whose work was consistently repetitive were 59% more likely to binge -

PAGE 29

23 drink than t hose whose work was repetitive none or some of the time; those whose work was repetitive most of the time were also 59% more likely. These associations were robust to adjustments for mental health variables. With respect to work hours, Table 3 shows that m others who worked over time (more than 40 hours a week) were 89% more likely to smoke than those who worked part time, a difference that was robust to adjustments for stress and depression. Two other work variables were significantly associated with an in creased likelihood of smoking: physical labor and night/evening shift work. Mothers whose work required moderate or hard physical labor were 33% more likely to smoke than those in seated jobs; this difference was attenuated with the addition of stress and depression variables. Similarly, mothers who worked regular evening or night shifts were 55% more likely to smoke than those who worked regular day shifts. This difference was robust to adjustments for stress and depression. None of the other work variables were significantly associated with smoking, binge drinking, or poor diet. Marital status and personal earning were also significantly associated with two of the outcomes. Married mothers were less likely to smoke and less likely to drink than single mothers; both of these relationships were robust to the addition of mental he alth variables. Marital status was not, however, associated with engagement in poor dietary habits. Personal earning was associated with smoking, with mothers who made less than ten thousand dollars at greater risk for smoking when compared to mothers whos e earning fell in to any of the other three personal income brackets. This relationship was robust to the addition of mental health variables. No other family variables were associated with smoking, binge drinking, or poor diet; however, stress scores were associated with a 6% higher likelihood of smoking, when controlling for all other variables.

PAGE 30

24 Work Family Characteristics and Biomarkers of Metabolic Dysfunction Results shown in Table 4 assess the relationship between work and family characteristics a nd three markers of early life metabolic dysfunction: BMI class, blood pressure class, and diabetic class. The health risk behaviors were included in all models; stress and depression scores were tested for any mediating effects. Of the work variables, onl y number of hours was associated significantly with any of the outcomes. When controlling for all other covariates, including stress and depression scores, compared to those who worked part time, those who worked full time were 51% more likely to be overwe ight or obese. Additionally, those who worked over time were 81% more likely to be overweight or obese. Compared to those who worked part time, those who worked over time were also 67% more likely to be pre hypertensive or hypertensive. This relationship w as robust with the inclusion of stress and depression variables. Work hours were not, however, significantly associated with being pre diabetic or diabetic. Among the family variables, marital status was associated only with BMI class: in the full model, married mothers were 35% more likely to be overweight or obese than single mothers. Personal earnings were also associated with this domain of outcomes: in the full model, compared to those who made ten thousand or less per year, those who made forty thous and or more were 43% less likely to be overweight or obese; 35% less likely to be hypertensive or pre hypertensive; and 38% less likely to be pre diabetic or diabetic. Further, those who made between thirty one thousand and forty thousand a year were 36% l ess likely to be pre diabetic or diabetic. None of the other work or family characteristics, nor the health behavior variables, were significantly associated with BMI class, blood pressure class, or diabetic class, and neither stress nor depression scores showed evidence of any mediating properties in the models.

PAGE 31

25 CHAPTER V DISCUSSION Extant sociological research around the intersection of work, family and health has primarily identified employment as beneficial for mothers, with social support and family circumstances either amplifying or reducing these benefits (Frech & Damaske, 2012; Hill, 2005). Using data from wave IV of the Adolescent to Adult Health (Add Health) survey, the current study builds on this literature by f ocusing on detailed aspects of employment and their association with not one, but three health domains for young, working mothers: mental health, behavioral health, and physiological health. In addition, this study takes into account the multifaceted nature of mental health by using measures of both stress and depression. As a result, this study yielded several important outcomes. First, I found that decision making latitude, hours worked and repetitiveness of work were significantly associated with self reported mental health. However, these occupational characteristics are not similarly associated with both subdomains of mental health : rather, hig her decision making latitude is associated with lower self reported stress, while repetitiveness of work is associated with higher self reported depression. Further, repetitiveness of work is significant in the behavioral health model, while decision makin g latitude is not. These differences suggest 1) the magnitude of the health benefits reaped by mothers who combine work and family may indeed be dependent upon the specific characteristics of said employment, wi th those in less favorable occupations receivi ng little or no health benefits , however 2) the benefit of certain occupational characteristics may heavily depend on what facet of mental health is being considered, with some characteristics associating with stress and others with depression.

PAGE 32

26 Second, there is evidence of class related disparities. Characteristics typically associated with working class jobs, such as physical labor, long hours, and night shifts, are associated with health risk behaviors and poorer physical health. However, these characteristics are not asso ciated with stress or depression scores. Explanations for this include 1) that these health behaviors function as coping behaviors, dampening experiences of psychological distress; 2) that these mothers have become somewhat desensitized to the strain of th eir work; or 3) engagement in these behaviors is tied to cultural norms associated with these forms of labor. There are several limitations to this study. First, the sample was fairly homogenous; mothers were overwhelmingly middle class, with approximatel y 12 children. Second, information about ethnicity was not available for this wave. T herefore, it is possible a great deal of the women who identified themselves as white were Hispanic white. Considering the cultural and socioeconomic differences between Hispanics, nonHispanic whites and nonHispanic blacks, as well as the large proportion of Hispanics in the United States, understanding differences between these three groups could lend greater nuance to this story. Third, we found no relationship between health behaviors and clinical biomarkers of health. This may suggest that our measures of health behaviors were suboptimal. A different operationalization of smoking, drinking and diet might yield different outcomes. Fourth, this study does not take in to account crucial elements of the life course, such as sequencing or timing. Age at first birth has profound consequences for mother’s long term health, to the extent that mothers who give birth before age 20 are at much higher risk for morbidity and mortality in mid age, even when controlling for early and later life SES (Henretta, 2007). Further, sequencing of marriage, employment and childbirth are no longer as linear as they may have once been. Women with more education are consistently delaying both chi ldbirth and marriage in favor of career advancement , while women with less

PAGE 33

27 education are delaying marriage, but not childbirth (Settersten, Fursetenburg, and Rumabut, 2008). Considering the cross sectional nature of this study, these nuances in timing and sequencing of family structure, and their influence on health, may be lost. Finally, detailed aspects of home life, such as whether there is a child with a disability at home or an unemployed spouse , are not considered in this study. By exploring the relationship between occupa tional characteristics and health for young working mothers, this study advances previous research and provides further insight into the ways in which labor experiences shape health outcomes for this population. The findings from this study also help outli ne future research trajectories. Specifically, longitudinal data should be collected to see how these relationships change over time; latent class analyses examining which clusters of characteristics are most important when discussing work, family and heal th for mothers is also crucial. Finally, an examination of how detailed aspects of home life (such as the presence of a child with a disability) either interact or cluster with employment circumstances to influence health would assist in painting a much m ore nuanced picture of how work and family combine to shape contemporary working mother’s well being and health.

PAGE 34

28 REFERENCES Acock, A. C. (2005). Working with missing values. Journal of Marriage and Family, 67(4), 10121028. doi:DOI 10.1111/j.17413737.2005.00191.x ACS. (2012). Working Mothers in the U.S. Retrieved from https://www.dol.gov/wb/Infographic_on_working_mothers.pdf Allen, T. D., & Armstrong, J. (2 006). Further examination of the link between work family conflict and physical health The role of health related behaviors. American Behavioral Scientist, 49(9), 12041221. doi:10.1177/0002764206286386 American Diabetes, A. (2015). (2) Classification an d diagnosis of diabetes. Diabetes Care, 38 Suppl , S8 S16. doi:10.2337/dc15S005 Benjamin, E. J., Blaha, M. J., Chiuve, S. E., Cushman, M., Das, S. R., Deo, R., . . . Stroke Statistics, S. (2017). Heart Disease and Stroke Statistics 2017 Update: A Report Fr om the American Heart Association. Circulation, 135(10), e146e603. doi:10.1161/CIR.0000000000000485 Berkman, L. F. (2015). Labor Policy and Work, Family, and Health in the Twenty First Century. Retrieved from http://www.ahrq.gov/professionals/education/ curriculum tools/population health/berkman.html Bianchi, S. M., Sayer, L. C., Milkie, M. A., & Robinson, J. P. (2012). Housework: Who Did, Does or Will Do It, and How Much Does It Matter? Social Forces, 91(1), 5563. doi:10.1093/sf/sos120 Buchmann, C., & M cDaniel, A. (2016). Motherhood and the Wages of Women in Professional Occupations. The Russell Sage Foundation Journal of the Social Sciences, 2 (4), 128150. CDC. (2016). Trends in Current Cigarette Smoking Among High School Students and Adults, United St ates, 19652014. Retrieved from https://www.cdc.gov/tobacco/data_statistics/tables/trends/cig_smoking/index.htm Chobanian, A. V., Bakris, G. L., Black, H. R., Cushman, W. C., Green, L. A., Izzo, J. L., Jr., . . . National High Blood Pressure Education Program Coordinating, C. (2003). The Seventh Report of the Joint National Committee on Prevention, Detection, Evalua tion, and Treatment of High Blood Pressure: the JNC 7 report. JAMA, 289(19), 25602572. doi:10.1001/jama.289.19.2560 Dallman, M. F., Pecoraro, N., Akana, S. F., la Fleur, S. E., Gomez, F., Houshyar, H., . . . Manalo, S. (2003). Chronic stress and obesity: A new view of "comfort food". Proceedings of the National Academy of Sciences of the United States of America, 100(20), 1169611701. doi:10.1073/pnas.1934666100 Damaske, S. (2011). For the Family? How Class and Gender Shape Women's Work : Oxford University Press Damaske, S. (2013). Work, Family, and Accounts of Mothers’ Lives Using Discourse to Navigate Intensive Mothering Ideals. Sociology Compass, 7(6), 436444.

PAGE 35

29 Damaske, S., Smyth, J. M., & Zawadzki, M. J. (2014). Has work replaced home as a haven? Re examining Arlie Hochschild's Time Bind proposition with objective stress data. Social Science & Medicine, 115 , 130138. doi:10.1016/j.socscimed.2014.04.047 DOL. (2012). Our History: An Overview 1920 2012. Retrieved from https://www.dol.gov/wb/info_about_wb/ interwb.htm DOL. (2014). Mothers and Families: Labor Force Participation. Retrieved from https://www.dol.gov/wb/stats/mother_families.htm Feeley, C., & Thomson, G. (2016). Tensions and conflicts in 'choice': Womens' experiences of freebirthing in the UK. Midwifery, 41, 1621. doi:10.1016/j.midw.2016.07.014 Frech, A., & Damaske, S. (2012). The relationships between mothers' work pathways and physical and mental health. J Health Soc Behav, 53 (4), 396412. doi:10.1177/0022146512453929 Frone, M. R., Barnes, G . M., & Farrell, M. P. (1994). Relationship of Work Family Conflict to Substance Use among Employed Mothers the Role of Negative Affect. Journal of Marriage and Family, 56(4), 10191030. doi:Doi 10.2307/353610 Frone, M. R., Russell, M., & Barnes, G. M. ( 1996). Work family conflict, gender, and health related outcomes: a study of employed parents in two community samples. J Occup Health Psychol, 1(1), 57 69. Goode, W. (1960). A Theory of Role Strain. American Sociological Review, 25(4), 483496. Han, W. J., Ruhm, C. J., Waldfogel, J., & Washbrook, E. (2008). The timing of mothers' employment after childbirth. Monthly Labor Review, 131(6), 1527. Hays, S. (1996). The Cultural Contradictions of Motherhood: Yale University Press. Henretta, JC. (2007). Early childbearing, marital status, and women's health and mortality after age 50. Journal of Health and Social Behavior, 48(3), 254266. Hill, E. J. (2005). Work family facilitation and conflict, working fathers and mothers, work family stressors and support. Journal of Family Issues, 26(6), 793819. doi:10.1177/0192513x05277542 Hochschild, A. (1997). The Time Bind: When Work Becomes Home and Home Becomes Work . New York, New York: Holt Paperbacks. Jackson, J. S., Knight, K. M., & Rafferty, J. A. (2010). Race and Unhealthy Behaviors: Chronic Stress, the HPA Axis, and Physical and Mental Health Disparities Over the Life Course. American Journal of Public Health, 100(5), 933939. doi:10.2105/Ajph.2008.143446 Jackson, S. L., King, S. M., Zhao, L., & Cogswell, M. E. (2016). Prevalence of Excess Sodium Intake in the United States NHANES, 20092012. MMWR Morb Mortal Wkly Rep, 64(52), 13931397. doi:10.15585/mmwr.mm6452a1 Kalleberg, A. (2011). Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s 2000s . New York, New York: Russel Sage Foundation.

PAGE 36

30 Karasek, R. A. (1979). Job Demands, Job Decision Latitude, and Mental Strain Implications for Job Redesign. Administrative Science Quarterly, 24(2), 285308. doi:Doi 10.2307/2392498 Kirschbaum, C., Wust, S., & Strasburger, C. J. (1992). Normal Cigarette Smoking Increases Free Cortisol in Habitual Smokers. Life Sciences, 50(6), 435442. doi:Doi 10.1016/00243205(92)903783 Klerman, J. A., & Leibo witz, A. (1999). Job continuity among new mothers. Demography, 36(2), 145155. doi:Doi 10.2307/2648104 Kochanek, K. D., Murphy, S., Xu, J., & Arias, E. (2017). Mortality in the United States, 2016. Retrieved from https://www.cdc.gov/nchs/data/databriefs/db 293.pdf Livingston, G. (2013). The Rise of Single Fathers . Retrieved from http://assets.pewresearch.org/wp content/uploads/sites/3/2013/07/single fathers 072013.pdf MacFarlane, L., & Montgomery, C. (2010). Influencing the mind through the body: A theoreti cal model for coping with student teacher stress through physical exercise. Journal of Cognitive Education and Psychology, 9, 183197. Marks, S. R. (1977). Multiple Roles and Role Strain Some Notes on Human Energy, Time and Commitment. American Sociological Review, 42(6), 921936. doi:Doi 10.2307/2094577 Marmot, M. G., Smith, G. D., Stansfeld, S., Patel, C., North, F., Head, J., . . . Feeney, A. (1991). Health inequalities among British civil servants: the Whitehall II study. Lancet, 337(8754), 13871393. Martin, M. A., Lippert, A. M., Chandler, K. D., & Lemmon, M. (2018). Does mothers' employment affect adolescents' weight and activity levels? Improving our empirical estimates. SSM Popul Health, 4, 291300. doi:10.1016/j.ssmph.2018.03.003 McDonald, P. ( 2000). Gender equity in theories of fertility transition. Population and Development Review, 26(3), 427+. doi:DOI 10.1111/j.17284457.2000.00427.x Menke, A., Casagrande, S., Geiss, L., & Cowie, C. C. (2015). Prevalence of and Trends in Diabetes Among Adults in the United States, 19882012. JAMA, 314(10), 10211029. doi:10.1001/jama.2015.10029 Mezuk, B., Ratliff, S., Concha, J. B., Abdou, C. M., Rafferty, J., Lee, H., & Jackson, J. S. (2017). Stress, self regulation, and context: Evidence from the Health and Retirement Survey. SSM Popul Health, 3, 455463. doi:10.1016/j.ssmph.2017.05.004 Miedema, M. D., Petrone, A., Shikany, J. M., Greenland, P., Lewis, C. E., Pletcher, M. J., . . . Djousse, L. (2015). Association of Fruit and Vegetable Consumption During Early Adulthood With the Prevalence of Coronary Artery Calcium After 20 Years of Follow Up: The Coronary Artery Risk Development in Young Adults (CARDIA) Study. Circulat ion, 132(21), 19901998. doi:10.1161/CIRCULATIONAHA.114.012562 Moen, P. & Chermack, K. (2005). Gender Disparities in Health: Strategic Selection, Careers and Cycles of Control. Journals of Gerontology, 60(B), 99 108. doi: 10.1093/geronb/60.Special_Issue_2. S99

PAGE 37

31 Musick, K., Meier, A., & Flood, S. (2016). How Parents Fare: Mothers' and Fathers' Subjective Well Being in Time with Children. American Sociological Review, 81(5), 10691095. doi:10.1177/0003122416663917 National Center for Chronic Disease Prevention and Health Promotion, D. o. C. P. a. C. (2013). Preventable Deaths from Heart Disease & Stroke. CDC Vital Signs. Retrieved from https://www.cdc.gov/vitalsigns/heartdisease stroke/index.html Nelson, C. C., Li, Y., Sorensen, G., & Berkman, L. F. (2012). As sessing the Relationship Between Work Family Conflict and Smoking. American Journal of Public Health, 102(9), 17671772. doi:10.2105/Ajph.2011.300413 Palumbo, A. J., De Roos, A. J., Cannuscio, C., Robinson, L., Mossey, J., Weitlauf, J., . . . Michael, Y. ( 2017). Work Characteristics Associated with Physical Functioning in Women. International Journal of Environmental Research and Public Health, 14(4). doi:ARTN 42410.3390/ijerph14040424 Park, C. L., & Iacocca, M. O. (2014). A stress and coping perspective on health behaviors: theoretical and methodological considerations. Anxiety, Stress and Coping, 27(2), 123127. Pearlin, L. I., Lieberman, M. A., Menaghan, E. G., & Mullan, J. T. (1981). The stress process. J Health Soc Behav, 22 (4), 337356. Peele, S., & Brodsky, A. (2000). Exploring psychological benefits associated with moderate alcohol use: a necessary corrective to assessments of drinking outcomes? Drug and Alcohol Dependence, 60(3), 221247. doi:Doi 10.1016/S03768716(00)001125 Robinson, L. D., Magee , C., & Caputi, P. (2016). Burnout and the work family interface A two wave study of sole and partnered working mothers. Career Development International, 21(1), 3144. doi:10.1108/Cdi 0620150085 Rod, N. H., Gronbaek, M., Schnohr, P., Prescott, E., & Kri stensen, T. S. (2009). Perceived stress as a risk factor for changes in health behaviour and cardiac risk profile: a longitudinal study. J Intern Med, 266 (5), 467475. doi:10.1111/j.13652796.2009.02124.x Romelsjo, A., Hasin, D., Hilton, M., Bostrom, G., D iderichsen, F., Haglund, B., . . . Svanstrom, L. (1992). The relationship between stressful working conditions and high alcohol consumption and severe alcohol problems in an urban general population. Br J Addict, 87(8), 11731183. Royston, P. (2004). Mult iple Imputation of Missing Values. Stata J, 4(3), 227241. Sabbath, E. L., Mejia Guevara, I., Noelke, C., & Berkman, L. F. (2015). The long term mortality impact of combined job strain and family circumstances: A life course analysis of working American m others. Social Science & Medicine, 146, 111119. doi:10.1016/j.socscimed.2015.10.024 Settersten, R.A., Furstenburg, F.F., Rumbaut, R.G. (2008). On the Frontier of Adulthood: Theory, Research, and Public Policy . Chicago, IL: The University of Chicago Press. Sieber, S. D. (1974). Toward a Theory of Role Accumulation. American Sociological Review, 39(4), 567578. doi:Doi 10.2307/2094422

PAGE 38

32 Suls, J., & Bunde, J. (2005). Anger, anxiety, and depression as risk factors for cardiovascular disease: the problems and implications of overlapping affective dispositions. Psychol Bull, 131(2), 260300. doi:10.1037/00332909.131.2.260 Sutherland, J. A. (2010). Mothering, Guilt and Shame. Sociology Compass, 4(5), 310321. Taris, T., Ybema, J. F., Beckers, D. G. J., Verheijden, M. W., Geurts, S. A. E., & Kompier, M. A. J. (2011). Investigating the Associations among Overtime Work, Health Behaviors, and Health: A Longitudinal Study among Full time Employees. International Journal of Behavioral Medicine, 18, 352360. Wall, G. (2001). Moral Constructions of Motherhood in Breastfeeding Discourse. Gender & Society, 15(4), 592 610. Wang, W., Parker, K., & Taylor, P. (2013). Mothers Are the Sole or Primary Provider in Four in Ten Households with Children; Public Conflicted about the Growing Trend . Retrieved from Pew Research Center: http://www.pewsocialtrends.org/2013/05/29/breadwinner moms/ Warren, G. W., Alberg, A. J., Kraft, A. S., & Cummings, K. M. (2014). The 2014 Surgeon General's r eport: "The health consequences of smoking -50 years of progress": a paradigm shift in cancer care. Cancer, 120(13), 19141916. doi:10.1002/cncr.28695 Wiesner, M., Windle, M., & Freeman, A. (2005). Work stress, substance use, and depression among young adu lt workers: an examination of main and moderator effect model. J Occup Health Psychol, 10(2), 8396. doi:10.1037/10768998.10.2.83 Yua, W. H., & Kuo, J. C. L. (2017). The Motherhood Wage Penalty by Work Conditions: How Do Occupational Characteristics Hinde r or Empower Mothers? American Sociological Review, 82(4), 744 769. doi:10.1177/0003122417712729

PAGE 39

33 Table 1. Descriptive Statistics (N=3,227)

PAGE 40

34 Table 2. Work, Family and Psychological Distress (N=3,227) *p<.05; **p<.01; ***p<.001 Note : Model 1 (not shown) includes only family variables and is not shown for parsimony’s sake. All models are adjusted for sex, age, race, education, household income and personal income.

PAGE 41

35 Table 3. Work, Family a nd Health risk behaviors (N=3,227) *p<.05; **p<.01; ***p<.001 Note: All models are adjusted for sex, age, race, education, household income and personal income. Family variables included in analyses but not shown in model.

PAGE 42

36 Table 4. Work, Family, and Car diometabolic Health (N=3,227) *p<.05; **p<.01; ***p<.001 Note : All models are adjusted for sex, age, race, education, household income and personal income. Family variables included in analyses but not shown in model.