ETHNICITY is a central determinant of health in the United States1 and represents a complex mixture of socioeconomic, cultural, and social structural influences on health. Findings of racial/ethnic disparities in health have highlighted important differences on key health outcomes or health indicators for women2 and are serving to mobilize research, policy, and programs to reduce and ultimately to eliminate disparities. Some have argued, though, that race, and presumably ethnicity, is a proxy measure for numerous factors including socioeconomic differences.3 For example, the recent longitudinal analysis of women's weight gain in the Alameda County Study showed that much of the excess weight gain of African American women compared with White women was explained by socioeconomic position during childhood and adulthood.4 Nonetheless, the influence of socioeconomic status on health in the United States has often received less attention in comparison to race and ethnicity because of lack of enthusiasm for programs aimed at distributing greater resources to disadvantaged populations.
Another important facet of ethnicity as a consideration in healthcare is sensitivity to ethnic differences in values, beliefs, and health practices of populations served. Awareness of such differences may facilitate tailoring programs so that these are more congruent with the values, beliefs, and health practices of populations served. The related concept of cultural competence is now an accepted part of education for the health professions. Incorporating a patient's cultural context into healthcare interactions and decision making is now indisputably recognized as important though often lacking in implementation.5
In this article, however, we pursue a less-traveled path in exploring ethnicity, as it relates to health, particularly among low-income women after childbirth. That path is through exploration of the dimensionality of postpartum thriving within 3 ethnic groups of low-income women: Anglo, African American, and Hispanic. By dimensionality we mean the organizational schema for defining the interplay of key postpartum variables, such as stress, social support, weight, and body image. Rather than reporting ethnic comparisons, the approach we take is a descriptive one, emphasizing the dimensionality of women's postpartum thriving within a slice of time, 3 months postpartum. Exploratory factor analysis of psychosocial and weight-related variables provides the vehicle for revealing the dimensionality of postpartum thriving within these 3 ethnic groups of low-income women. Our goal is to reveal the unique organization of psychosocial space around concepts such as stress, social support, and body image within ethnic groups.
Furthermore, it is our belief that there may be more "spill over" among life domains among vulnerable populations, such as low-income women, requiring more integrative perspectives.6 That is, being able to segment or compartmentalize roles and experiences from one life domain to another requires energy and resources that may be less available among vulnerable populations where available energies are focused on more immediate needs. Thus, despite contexts that may differ as a result of ethnicity, low-income women may share the press of economic disadvantage and its consequences in how they organize their experience during life transitions, such as the postpartum period. Knowledge of the dimensionality of thriving in low-income, ethnically diverse women may contribute to more tailored and relevant approaches to health promotion during the postpartum period. To develop these ideas, we first review the construct of thriving and a factor analysis of thriving-related variables in an advantaged sample of postpartum women. Following this, we present findings on the dimensions of thriving among low-income Anglo, African American, and Hispanic women.
BACKGROUND
The construct of thriving
The impetus for this article comes from an earlier exploration of the dimensionality of thriving in a relatively advantaged sample of new mothers studied by survey methods at 3 and 12 months postpartum reported by Walker and Grobe.7 That work identified the dimensions of thriving among women who had given birth and were consequently undergoing the multiple physical, psychological, and social changes that accompany new motherhood. In recognizing the complexity of these changes, Walker and Grobe proposed the construct of thriving as an integrative concept to capture the psychosocial and weight-related dimensions of health and well-being during the postpartum transition. Thriving was defined as "the dynamic relationship between lifestyle, psychosocial factors, and nutrition (reflected in weight status) during the postpartum period that enhances the well-being of mother and, if she is lactating, the infant."7(p153)
In formulating the construct of thriving in the childbearing context, Walker and Grobe drew on 2 conceptual traditions: the work of Orem8 related to self-care, and the long history of clinical work related to failure to thrive that was later reformulated as the positive concept of thriving by a faculty group at the University of North Carolina at Greensboro.7
Thriving was articulated conceptually within Dorothea Orem's self-care theory,8 which stresses the essential process of caring for the self (vs neglect) in the context of life stage, and "self-care agency" as the capability to meet developmental life-stage requirements. Thus, self-care is defined as "the personal care that individuals require each day to regulate their own functioning and development."8(p8) During the childbearing years, women have life stage-related "self-care requisites" for promoting and maintaining their health, well-being, and functioning. These healthcare requisites or needs of women during pregnancy and the postpartum period generally fall into a class in which "the health focus is oriented to the life cycle" and "promoting and maintaining health and [horizontal ellipsis] protecting against specific diseases and injuries."8(p135) Women meet the needs of this life stage by drawing on their own inner resources, as well as seeking needed knowledge, skills, and support from their partners, families, communities, and the healthcare system. In Orem's theory,8 ethnicity is a basic conditioning factor that, along with factors such as age and socioeconomic status, shapes a person's self-care agency. Orem's framework, although sometimes misunderstood as advocating personal independence over interdependence, is useful in recognizing the importance of developmental stages in healthcare, and the numerous resources that positively and negatively influence health. Still, it is a broad framework, and lacks the specificity needed to elaborate the key requisites and challenges at specific life stages and for persons in specific cultural and socioeconomic contexts.
For this reason, Walker and Grobe7 also drew on work related to thriving for its relevance to life-stage requisites and functioning. Thriving as a clinical concept has its origins in clinical descriptions of inorganic failure to thrive, a phenomenon first observed among children9 and then among elders.10 A hallmark of the manifestations in each of these groups was a blending of a wasting nutritional state evidenced in severe underweight or weight loss with accompanying psychosocial alterations. That is, severely impaired nutritional status in inorganic failure to thrive was a complex multilevel phenomenon, and distinct from those cases of weight loss stemming from organic causes in children and adults. Subsequently, the work of the Thriving Work-Group at the University of North Carolina, Greensboro, was instrumental in redefining failure to thrive as thriving, a positive concept of lifespan relevance.7 Walker and Grobe further proposed that thriving might be viewed as spanning a continuum of high to low levels that aided in understanding developmental needs of childbearing women related to "manag[ing] nutritional, psychosocial, and lifestyle concerns."7(p152)
The dimensionality of thriving among advantaged women
Building on their proposal that nutritional (including weight), psychosocial, and lifestyle concerns were ingredients of postpartum thriving, Walker and Grobe7 used factor analysis to empirically examine the dimensions of postpartum thriving, which was conducted on an existing dataset collected from 145 women during the first postpartum year. Most of the women had some college education (83%) and most women were either Anglo (77%) or Hispanic (17%).
The outcome of the factor analysis, which included varimax rotation of factors, was 4 orthogonal factors:
* Psychosocial distress, defined by depressive symptoms, stressful events, and (low) social support;
* Lifestyle patterns, defined by overall health practices, physical activity, and dietary habits;
* Weight, defined by body mass index and weight gain since prepregnancy; and
* Body image, defined by body image dissatisfaction and emotional self-regulation.
These findings suggested that within this relatively advantaged sample, the areas of psychosocial distress, lifestyle patterns, weight, and body image represented somewhat distinct and independent aspects of postpartum women's lives.
One shortcoming of the analysis was that the variables were measured at different times, with some measured about 3 months and others about 12 months postpartum. One of the key findings of this study, though, was that psychosocial distress was the first and most strongly defined factor in the analysis, and this definition (based on factor loadings of 0.79 for depressive symptoms, 0.65 for stressful events, and -0.51 for social support) delineated the manifestation of that distress: elevated depressive symptoms, high stress, and low social support. Conversely, psychosocial thriving would be manifest in low depressive symptoms, low stress, and high social support. What this analysis left uncertain was what the structure of thriving would look like in less advantaged women. For example, low-income women might lack resources that support a healthy lifestyle and contribute to reduced depression. Thus, the "phenomenology" of various thriving factors might look different in low-income women. Furthermore, it might also vary on the basis of new mothers' ethnicity to the extent that ethnicity shaped psychosocial and related aspects of their daily living.
METHODS AND RESULTS: DIMENSIONS OF THRIVING AMONG LOW-INCOME WOMEN
To explore the dimensionality of thriving among low-income women, a factor analysis is reported here that makes use of data from the Austin New Mothers Study (ANMS). The ANMS was a prospective study of weight and its psychosocial context among Anglo, African American, and Hispanic women during the first 12 months postpartum. A detailed description of this sample of low-income women is published elsewhere11,12 and their characteristics will be described only briefly here. The analyses reported here are based on 358 participants (106 Anglo, 90 African American, and 162 Hispanic women), with complete data on 9 key variables (Table 1) at 3 months postpartum. All participants had received perinatal care funded by Medicaid and were at 185% or less of the federal poverty level. The mean age of women whose data are reported was 22.3 years (SD = 3.9), their mean parity was 1.9 (SD = 0.8), and their mean body mass index was 28.6 (SD = 6.4) at 3 months postpartum. Most women were either high school graduates (38%) or had a partial high school education (39%). Psychometric scales had evidence of reliability and validity.11,12
The 9 psychometric scales and other measures (Table 1) included in this factor analytic study, conducted with varimax rotation, are very similar to those in the earlier factor analysis reported by Walker and Grobe,7 with one exception. To maintain a minimum variables-to-subjects ratio of 4:1, we omitted one variable, weight gain (referenced to prepregnant weight), from the analysis. Maintaining this ratio was especially important in the analysis involving African American women who composed the smallest number of cases (n = 90) among the 3 ethnic groups.
Table 2 presents the factor dimensions with factor loadings for each of the 3 ethnic groups. Within each ethnic group, 3 factors were extracted. The first and most robust factor was psychosocial distress manifested in loadings of stress, depressive symptoms, social support (negative loading), (less healthy) lifestyle, and (less favorable) body image. (Note: Where the directionality of a loading may be unclear to readers unfamiliar with a scale, we have inserted a directional comment parenthetically.) For Anglo (n = 106) and Hispanic (n = 162) women, the variable of self-regulation (ie, emotional self-management) also had a negative loading on the distress factor. The second and third factors for Anglo and Hispanic women-lifestyle and weight-were generally defined by weaker factor loadings. The second and third factors for African American women (n = 90), lifestyle and regulation, were defined by single variables with high loadings. In general, the factors extracted for Anglo and Hispanic women were similar. Those for African American women, especially on their third (regulation) factor, were unique to that group. Finally, several variables had high loadings on more than one factor. For example, for both Anglo and Hispanic women, the variables of body image and emotional regulation loaded on the psychosocial distress factor as well as the weight and lifestyle factors, respectively.
DISCUSSION
The dimensionality of thriving
In contrast to factor analysis of postpartum variables of more advantaged women that resulted in extraction of the 4 clearly defined factors of psychosocial distress, lifestyle patterns, weight, and body image,7 factor analysis of similar variables among low-income women revealed a smaller set of 3 factors: psychosocial distress, lifestyle, and weight for Anglo and Hispanic women; psychosocial distress, lifestyle, and regulation for African American women. This may be due in part to the deletion of one weight-related variable from the analysis. For the most part, though, the reduction in the number of factors extracted was related to incorporation of variables from the body image factor in the earlier factor analysis into the psychosocial distress or other factors in this analysis among low-income women.
Of particular interest was the increased number of variables that loaded on the psychosocial distress factor among low-income women. In the earlier analysis of more advantaged women, psychosocial distress was defined by high loading of depressive symptoms, stressful events, and social support (negative loading). These variables had high loading in each ethnic group of low-income women examined here, but were also accompanied by loadings of more than 0.40 for overall health-related lifestyle (higher score is riskier) and body image (higher score is more dissatisfied). This broadened phenomenology of psychosocial distress is consistent with our view that low-income women may have fewer reserves to compartmentalize life domains, which in turn result in "spill over" among domains. Consequently, effects of stress or other life disruptions are more likely to cut across multiple life domains, such as feelings of depression, health behaviors, and perceptions of self. It is also possible that the larger number of variables loading on the psychosocial distress factor could have arisen for other reasons. These include method variance and the tendency of depressed persons to make negative attributions.24,25 Although we cannot eliminate these alternative hypotheses, we note that one scale, the Food Habits Questionnaire, which incorporated methods similar to scales loading on the psychosocial distress factor, had very low loadings on that factor.
At a more conceptual level, several observations regarding the findings are noteworthy. First, viewed from the perspective of Orem,8 the management of psychosocial distress appears to be a key self-care requisite among low-income postpartum women, regardless of ethnicity. However, the experience of psychosocial distress is not limited to the emotional realm but spills over into functional self-care behaviors reflected in health-related lifestyle. Unlike the earlier classical findings related to failure to thrive, however, postpartum weight and psychosocial functioning were not closely interwoven, at least as evidenced by factor loading values. Finally, although emotional self-regulation captures only a portion of the meaning of Orem's concept of self-care agency, it is noteworthy that the scale tapping self-regulation loaded on both the psychosocial distress and lifestyle factors for Anglo and Hispanic women but constituted a unique factor for African American women. Self-care agency is clearly relevant to managing psychosocial distress and lifestyle, thus the dual loading of emotional self-regulation among Anglo and Hispanic women is reasonable. It is less clear why emotional self-regulation is a factor unto itself among African American women. It may reflect coping strategies developed by African American women that lead them to develop self-regulatory responses that are independent of other factors such as stress and health-related lifestyle.
Considerations for intervention
From the standpoint of the phenomenology of psychosocial distress among low-income postpartum women, the findings reported here reveal that psychosocial distress is a very encompassing experience manifested not only by depressive symptoms and stress in the context of low social support but also by riskier health practices, dissatisfaction with one's body, and, for Anglo and Hispanic women, their perceived ability to self-regulate emotions and thoughts. (For Anglo and Hispanic women, the variables of body image and self-regulation also loaded on weight and lifestyle factors, respectively.) These factor analytic findings suggest that interventions to ameliorate psychosocial distress among low-income women, regardless of ethnicity, potentially may need to consider including a range of intervention content, perhaps including social cognitive approaches, building coping skills, health education, and relationship and social support resources. Recent research has documented the benefit of a social cognitive intervention in reducing the self-regulatory behavior of negative thinking and the incidence of depressive symptoms in low-income single African American and White mothers determined to be at risk for depression.26 The intervention consisted of a 4- or 6-week series of individual and small group meetings in which strategies to manage negative thinking were taught and practiced. Participants then were expected to practice these techniques between weekly sessions to further develop this personal skill. Longitudinal analyses at 6 months postintervention documented a statistically significant decrease in the incidence of depressive symptoms, occurrence of perceived life stressors, and fewer negative thoughts during the postintervention period with the sample of low-income mothers.
Further enrichment of psychosocial programs with health-related lifestyle content, such as nutrition, physical activity, and weight management, might have additive effects and also would address other dimensions of the postpartum experience that may be relevant to low-income women in many communities. Evidenced-based models for such content include the "Moms on the Move" intervention,27 which focused on increasing the physical activity of low-income women participating in the Special Supplemental Food Program for Women, Infants, and Children. Through a combination of provider counseling and interactive brochures, this 8-week program resulted in significant increases in mothers' physical activity as well as activity-related constructs, such as self-efficacy for exercise.
In conclusion, the postpartum period is a critical lifespan developmental phase in numerous challenges to self-care and health. The factor analysis of thriving/distress described here is one approach in identifying culturally and socially relevant components of interventions to maximize positive self-care behaviors and thriving during the postpartum period. By helping new mothers to identify their specific challenges and then by providing instruction, social support, and a safe, secure context in which to practice new cognitive and behavioral skills, family and community health professionals can provide an integrative perspective and an environment for postpartum thriving to be realized.
Limitations and strengths
Our findings are based on one sample of low-income women at one time point during the postpartum period. Our selection of the 3-month time period was based on the availability of variables of interest in the ANMS dataset and the time point that had the most favorable variables-to-subjects ratio. The 3-month time point also corresponds to the measurement of a portion of variables in the earlier factor analysis of Walker and Grobe.7 Given the study measures, however, it was not possible to determine whether any women were experiencing a major postpartum depression versus psychosocial distress stemming from contextual influences.
Qualitative research methods are often seen as the method of choice for extracting themes defining a person's experience of a transition. Employing factor analysis to explore dimensionality of a life-stage transition is a less frequently used alternative method. This report explored the dimensionality of thriving in low-income women, particularly the phenomenology of postpartum psychosocial distress.
REFERENCES