Community Prevention Research in Women’s Health

The Women’s Health Initiative is a project of the National Institutes of Health, and since 1995, the Prevention Research Centers have conducted research for the initiative’s community prevention component. Results from this research were shared at the Community Prevention Research in Women’s Health Conference, October 26-27, 2000, at the National Institutes of Health, Bethesda, Maryland. Abstracts of the presentations made by PRC researchers are collected here.

Reducing the Risk of Cardiovascular Disease Among Black Women

Author:
Diane M. Becker, ScD, MPH;
Lisa R. Yanek, MPH;
Taryn F. Moy, MS, RD;
Joel Gittelsohn;
Dyann Matson Koffman, DrPH, MPH, CHES

Background

Urban African American women aged 40 years or older have a marked excess risk of obesity, sedentary lifestyle, and cardiovascular disease (CVD) morbidity and mortality. Public health strategies to lower the risk of lifestyle-related CVD among women in this population have met with little long-term success. We used three different church-based nutrition and physical activity strategies to examine the impact on the CVD risk profiles of women in a single year.

Methods

The strategies were standard behavioral group interventions, behavioral and spiritual strategies, and self-help interventions. Women were screened at baseline and at one year. We used the generalized estimating equations adjustments for intrachurch clustering to determine intention-to-treat within each group and between groups. Because no differences existed between the two behavioral groups, we pooled them for comparison with the self-help group.

Results

Our sample consisted of 529 women (mean age 53 ± 9 years) from 16 churches in Baltimore. We observed statistically significant differences in risk factor profiles between the behavioral intervention group and the self-help group. Body weight (-1.1 lbs), waist circumference (-0.66 inches), systolic blood pressure (-1.6 mmHg), dietary energy (-114 kcal), dietary total fat (-8 g), and sodium intake (-141 mg) improved significantly among the intervention participants but not among women in the self-help group. At one year, larger clinically meaningful changes in risk factors occurred among women in the intervention group in the upper decile for weight loss (-19.8 lbs). A much smaller weight loss (-7 lbs) was observed in the upper decile of the self-help group, and changes in all other outcomes were small.

Conclusions

One year after the program was initiated, one in ten women in the intervention group had clinically important improvements in CVD risk profiles, but this improvement was not seen among women in the self-help group. If applied nationally, these findings suggest a public health benefit of church interventions for African American women.

---

 

Peer Support Intervention for Cardiovascular Disease Risk Among African American Women

Author:
Carol E. Cornell, PhD

Background

Significant racial disparities in morbidity and mortality from cardiovascular disease (CVD) exist for U.S. women. Minority women who live in rural communities where access to health care resources is limited are at particularly high risk. Risk-reduction interventions are often unavailable for these women because of limited funding, geographic inaccessibility, and other constraints. The Community Health Advisor (CHA) Model recruits and trains women as “natural helpers,” to provide resources for health care delivery and other programs in underserved areas.

Methods

Using a cross-sectional, quasiexperimental design, the Uniontown Community Health Project developed, implemented, and evaluated CHA-based behavioral interventions aimed at reducing CVD risk among African American women age 40 or older in Uniontown, Alabama. (York, Alabama, served as a comparison community.) Phase I included formative assessment, baseline community surveys, Community Health Council development, and identification and training of CHAs. Phase II consisted of a 12-month intensive intervention period, in which CVD risk-reduction classes and other activities (e.g., walking clubs, health fairs, smoke-free days, and blood pressure screenings) were co-conducted by CHAs and project staff. Environmental interventions were initiated, including clean-up and lighting of an outdoor walking track, renovation of a physical activity room, and increased availability of low-fat foods in local stores. During Phase III, the maintenance phase, CHAs conducted all risk-reduction activities, and project staff served as consultants and evaluators.

Results

Twenty-five persons completed general CHA training. Participation in classes totaled 388, while 590 people participated in community events where attendance was taken (such as health fairs, fund walks, and food fairs). A total of 606 women participated in baseline surveys, and 500 women were available for re-interview one year later in both Uniontown and York.
Preliminary results indicate 85 percent of African American women surveyed in Uniontown were aware of the project. Participants had decreased levels of stress, increased knowledge of CVD, increased physical activity, decreased cigarette smoking, and increased use of low-fat cooking techniques (all p values <.01.) Analyses are under way for the follow-up survey conducted at the maintenance period.

Conclusions

During Phase III, project staff worked with CHAs to develop methods for institutionalizing and maintaining risk-reduction activities in the community. These efforts have included partnerships with the city of Uniontown, the Alabama Extension Service, and the Alabama Department of Public Health. Process data indicate increased community capacity, as illustrated by the generation of new research projects and nonresearch activities in collaboration with the University of Alabama at Birmingham and other partners.

---

 

Osteoporosis Prevention: Education and Outreach

Author:
Susan J. Blalock, PhD;
Brenda M. DeVellis, PhD

Background

This study aimed to increase the percentage of women consuming recommended levels of calcium, to increase the percentage of women regularly engaging in light to moderate weight-bearing physical activity for at least 30 minutes per day, and to enhance women’s ability to make informed decisions about the use of hormone replacement therapy (HRT).

Methods

The study in 12 counties in western North Carolina used a factorial research design to evaluate the effects of two separate interventions. For the first intervention, conducted at the county level, we divided 12 participating counties into matched pairs on the basis of total population, percentage of whites in the population, and percentage of persons living in poverty. We then randomly assigned one county in each pair to the intervention and the other to serve as the control. The second intervention consisted of individual patient education, and we randomly assigned participants to tailored or nontailored treatment conditions. The patient education interventions in each county were evaluated independently from the countywide interventions. This allowed us to examine both main effects and interaction effects. Telephone interviews were used to collect data before implementation and 3, 6, and 12 months after implementation. We recruited 714 women aged 40 to 56 years; complete data through the 6-month follow-up interviews were available for 589 women. In analyses, we also stratified the sample by participants’ baseline stage of change and examined the effect of the patient education intervention among women in different stages. We examined the separate effects of the two interventions on calcium intake and exercise level at 3- and 6-months after baseline by using regression analysis.

Results

The community intervention had no effect on calcium intake, and we observed no interaction between the two interventions. However, we found a significant interaction between the tailored education intervention and participants’ stage of change (p <.001). The tailored intervention was superior to the nontailored one except among women in the earliest stages of change. Neither interventions had an effect on exercise behavior, either along or in combination.

Conclusion

Although the effects of the community intervention appear quite limited, broader effects were not captured in our formal evaluation. For example, we were much more successful in recruiting participants in the intervention than in the control counties. The community health forums were attended by persons who did not participate in our study, and in at least two communities, local health care providers began providing bone density testing following their involvement with this study. Further, 13 percent of the women recruited either had osteoporosis or were at very high risk of developing osteoporosis, as indicated by low bone density scores. The benefits associated with early detection for these women are not captured by our formal evaluation.

---

 

Assessment of Moderate Physical Activity of Minority Women in Texas

Author:
Louise C. Masse, PhD

Background

Most physical activity instruments have not been developed specifically to assess the lifestyle and health experiences of minority women. The Women on the Move study developed and validated physical activity instruments for African American and Hispanic women aged 40 years and older, and for measuring moderate and intermittent patterns of physical activity.

Methods

All instruments were validated against physical activity diaries, activity monitors (accelerometer), and the doubly labeled water methodology. We developed five different surveys: 1) a global survey, 2) a self-administered activity checklist, 3) a modified 7-day recall that assessed all types of physical activity, 4) an interviewer-administered survey that captured patterns of physical activity by type (PACT), and 5) a survey of physical activity recommendations. All surveys were translated into Spanish, translated back into English, and then refined to develop equivalent forms. To examine the validity of the surveys, we correlated the physical activity dimensions of the surveys against a 7-day physical activity diary and a 7-day accelerometer recording.

Results

A total of 130 African American and 130 Hispanic women participated in the study (mean age = 49.1 ± 7.0 yrs, mean BMI =30.0 ± 6.3 kg/m2). Correlations with the accelerometer were .20 for PACT, .33 for the checklist, and .36 for the global survey. Correlations with the physical activity dimensions involving all activities were .60 for PACT, .66 for the checklist and .34 for global survey. The surveys had adequate correlations with individual items in the physical activity diary, except for family care, transportation, and volunteer work. In general, the low correlations were associated with dimensions that had little variability.

Conclusion

Overall, the checklist performed better than the other two surveys and can be used for comprehensive assessment of physical activity. Thus, physical activities associated with occupation, home, exercise, and all activities can be validly measured and without any significant bias.

---

 

Assessment of Moderate Physical Activity of Minority Women in South Carolina

Author:
Barbara E. Ainsworth, PhD, MPH

Background

Many early studies of leisure-time physical activity recruited men only, and investigators developed questionnaires that focused on sports, conditioning, and lawn and garden activities. Women spent considerable time cleaning the house, caring for family members, and working in one or more paying jobs. New instruments are needed to capture activities performed by women, and the questionnaires should account for cultural and ethnic differences in physical activity. Through the Cross-Cultural Activity Participation Study (CAPS), we developed a culturally relevant questionnaire that measures the types of moderate and vigorous intensity physical activities performed by minority women aged 40 years and older.

Method

In this 5-year study, we included American Indian women living in pueblos and the Navajo Nation in New Mexico, African American women living in the midlands of South Carolina, and American Indian women living on the Leech Lake Indian Reservations in northern Minnesota. In Phase I, we observed the physical activity patterns of 200 women in New Mexico and South Carolina for 6 months. We obtained detailed information about each woman’s activity from physical activity records she kept for 12 days and from a motion sensor she wore. We also completed 50 in-depth interviews with American Indian and African American women to identify the role of physical activity in their lives. In Phase II, we used data from Phase I to develop and validate a questionnaire, the Typical Week Physical Activity Survey (TWPAS). To validate TWPAS, we used measures of physical activity (records, surveys, motion sensors), fitness (treadmill tests), measures of body fat, and heart disease risk factors (blood pressure and blood level of glucose, lipids, and insulin). In Phase III we further evaluated TWPAS to simplify and identify the most precise way to administer the questionnaire. We also measured the energy cost of three uniquely American Indian physical activities–jingle dancing, ricing, and building a sweat lodge.

Results

The TWPAS had acceptable validity (r=.51) for moderate intensity activities when compared with physical activity records. Vigorous intensity activities were related with higher fitness levels (r=.45). Individual items on the TWPAS had a correlation of .40 to .76 when compared with similar items in a physical activity record for housework, lawn and garden work, nonoccupational walking, dancing and sports, conditioning activities, leisure-time inactivity, occupation, and volunteer activities. Test-retest intraclass coefficients were from .55 to .75. In-depth interviews showed that all women found it difficult to schedule time for structured activities and that family support was very important to being physically active. American Indian women believed that cultural traditions around physical activity were important and that walking and dancing were activities that integrated their culture, family, tradition, and self-care.

Conclusion

Measuring only leisure-time sports and conditioning activities on a questionnaire underestimates women’s true level of physical activity. Asking women about the types of activities they consider to be important also validates and affirms their efforts and may increase the quality of data collected in studies of physical activity and health.

---

 

Environmental and Policy Interventions to Increase Physical Activity Among Minority Women

Author:
Ross C. Brownson, PhD;
Amy A. Eyler, PhD;
Abby C. King, PhD;
David R. Brown, PhD;
Yuh-Ling Shyu, PhD;
James F. Sallis, PhD

Background

Large national surveys have shown that women are less likely than men to be physically active and that women of some ethnic and racial groups may be at especially high risk for physical inactivity. We studied the patterns of physical activity among minority women by using various definitions to determine sociodemographic and behavioral correlates of physical activity.

Methods

We conducted a cross-sectional telephone survey of African American, American Indian or Alaskan Native, Hispanic, and white women (n= 2,912) aged 40 years and older. We did not include Asian American women because the response rate from women in this group was extremely low.

Results

Physical activity was lowest among African American and American Indian or Alaskan Native women. The adjusted odds ratio (OR) for no leisure-time activity was 1.35 for African American women and 1.65 for American Indian or Alaskan Native women. When we used occupational activity rather than traditional assessments of leisure activity, the proportion of women classified as physically active was much higher. Using a composite definition of physical activity, we found that 72 percent of respondents reported being physically active. Women who lived in rural regions (OR=1.33) were more likely to be completely inactive in leisure time than were women who lived in urban areas.

Conclusion

Not all groups were less active than white women when all domains of physical activity were taken into account, but minority women were generally less active.

---

 

Improving the Delivery of Diabetes Care to Women in Minority Groups

Author:
Thomas C. Keyserling, MD, MPH;
Carmen Samuel-Hodge, PhD, RD

Background

The prevalence of type 2 diabetes is almost twice as high for African American women (10.8%) than for white women (5.9%); the highest rates occur among older African American women. African American women who have diabetes are much more likely to have complications, including blindness, amputation, and end-stage renal disease, than are white women. Effective education programs that enhance self-care and access to high-quality medical care are needed to reduce diabetes-related morbidity and mortality.

Purpose

We developed and tested a comprehensive diet and physical activity behavioral intervention for African American women with type 2 diabetes. We designed the intervention to improve moderate intensity physical activity; dietary intake; other self-care behaviors such as foot care and glucose monitoring; and psychosocial factors, including perceived competence, self-efficacy, and quality of life.

Methods

We used focus groups in a sequential fashion to develop and refine the intervention. For the development phase, we convened 11 focus groups (n=76) to address physical activity, diet, psychosocial issues, and the role of a community advisor. To develop and test prototype materials, we used 6 focus groups (n= 47), cognitive response testing (n=26), and telephone interviews (n=45). The intervention included four sessions of individually tailored clinic-based counseling by a nutritionist and three group sessions of community-based counseling and monthly phone calls from a Community Diabetes Advisor (CDA)–a layperson with diabetes who served as a resource. We used a randomized controlled trial with three treatment groups:
Group A–clinic and community intervention, Group B–clinic intervention only, and Group C–minimal intervention (standard educational pamphlets). African American women aged 40 years and older with type 2 diabetes were recruited from seven primary care practices. Participants wore a Caltrack accelerometer for 7 days to measure physical activity. The Minnesota Nutrition Data System (NDS) was used to assess dietary intake. We used a variety of instruments to assess psychosocial outcomes, some of which were validated in previous studies.

Results

The average age of the 200 participants was 59 years, and the average length of time the women had had diabetes was 10.4 years. The mean glycosylated hemoglobin was 11.1, and the mean BMI was 36; 75 percent of the women had high blood pressure. The mean physical activity kilocalories per day (kcal/d) for Group A slightly decreased from 342 (SE=20.5, n=59) at baseline to 339 (SE=22.8, n=53) at 12 months. The mean physical activity kcal/d for Group B increased from 336 (SE=25.1, n=60) at baseline to 344 (SE=25.4, n=58) at 6 months and then decreased to 322 (SE=26.9, n=56) at 12 months. For Group C there was a substantial decrease in the physical activity kcal/d from 321 (SE=18.7, n=62) at baseline to 289 (SE=21.4, n=58) at 6 months before increasing to 297 (SE=22.0, n=56) at 12 months.

Conclusion

The intervention was feasible and acceptable and was associated with a modest but statistically significant increase in physical activity. The increase in physical activity for the clinic and community intervention (Group A) was significantly higher than that for the minimal intervention (Group C), but the increase for the clinic intervention only (Group B) was not statistically higher than that for the minimal intervention. This finding shows the importance of incorporating aspects of community intervention in clinic-based physical activity programs. The impact of the intervention on dietary change was not significant, but this finding may be in part due to the inadequacy of the 24-hour food recall methodology used.

---

 

Overview of the ENDOW Study: Ethnicity, Needs, and Decisions of Women

Author:
Cora E. Lewis, MD, MSPH

Background

Health issues during menopause, such as hysterectomy and use of hormone replacement therapy (HRT), may vary among women by culture and ethnicity. Yet, previous research on hysterectomy and HRT use have focused on medical issues. Few investigators have explored decision making, preferences, and values of women as well as differences by ethnicity. We explored ethnic, geographic, and cultural variations in attitudes and decision making concerning hysterectomy and use of HRT. We also developed and pilot tested materials appropriate for different sociocultural groups to enhance women’s decision-making skills.

Methods

The ENDOW study was a 5-year, three-phase, multicenter collaboration for studying women in diverse communities of Alabama, South Carolina, Texas, and New Mexico. We collected qualitative data from 78 focus groups and 480 women: African American (32%), Hispanic (32%), non-Hispanic white (23%), and Navajo (12%). We then used information from previous research, theory, and focus groups to develop a survey. This survey included a core instrument administered at all sites and additional items for specific populations. We conducted in-person interviews or administered the survey by telephone. We used the data from focus groups and surveys to develop intervention methods and strategies to facilitate women’s decision making. To enhance women’s decision-making skills, we developed materials, including a video, brochures, and a decision-making guide, for group or individual use. We then pilot tested the materials and the dissemination strategies.

Results

Respondents were about 3,500 women: African American (30%), Hispanic (19%), Navajo (11%), and non-Hispanic white (39%). The qualitative and survey data were consistent and showed that women want to be involved in medical decisions, want better information about hysterectomy and other health topics, and want improved communication with doctors.

Conclusion

We recommend four steps to approaching interventions, particularly those involving decision making, for menopause. 1) Design intervention tools for specific groups, including patients, providers, and health systems. 2) Formalize testing of decision-making interventions in controlled trials and focus on multiple outcomes. 3) Continue to refine intervention materials and tools for evaluations. 4) Develop effective dissemination strategies.

---

 

The Role of Qualitative Data in the ENDOW Study

Author:
Carla J. Herman, MD, MPH

Background

Focus groups, based on principles of marketing and psychology, are used to collect qualitative data on subjective values. We used this method to generate hypotheses about the experiences, perceptions, and attitudes of a multicultural population of women concerning menopause, hormone replacement therapy (HRT), and hysterectomy.

Methods

Each focus group session was taped, transcribed, and translated by native speakers; each transcript was coded for thematic analyses by ethnicity and culture. The University of Alabama conducted five focus groups with 38 women aged 30 to 76 years who had had a hysterectomy in the past 3 years. The University of New Mexico conducted 23 focus groups with 156 women aged 34 to 71 years who were recruited in five communities and the eastern Navajo Nation. The groups were homogeneous for hysterectomy status, language, and ethnicity. The University of South Carolina conducted 17 focus groups with 82 women aged 30 to 65 years recruited from two coastal communities. These women were homogeneous for hysterectomy status and ethnicity. The University of Texas at Houston conducted 23 focus groups with 148 women (mean age about 44 years). The participants in Texas were homogeneous for language and ethnicity, and none of the women had had a hysterectomy.

Results

Regardless of cultural and ethnic group, women wanted more information about hysterectomy and wanted to feel involved in the decision-making process. Although the women reported overall positive outcomes for hysterectomy, they felt unprepared for the recovery process. Some participants reported a lack of good communication with their physicians. Distrust toward physicians was higher among African American women than those of other ethnic groups. Compared with women of other groups, African American and Hispanic women believed men’s views of hysterectomy were negative, and men’s perceptions were more important to Hispanic women.

Conclusion

Women want to be involved in making decisions about their health, and they want to involve family members as well. Women want to thoroughly discuss surgery and medical management as well as alternatives, and to learn what to expect during recovery. Cultural and spiritual aspects of decision making should be included in the educational process.

---

 

Preliminary Analysis of Mental and Physical Health Outcomes of Hysterectomy Experiences

Author:
Robert E. McKeown, PhD

Background

Hysterectomy is the most common surgery among women aged 45 to 64 years; 89 percent of women elect to have this surgery. We explored whether physical and mental health scores were higher for women who had never considered having a hysterectomy than for women who had. We also compared women who had or had not had a hysterectomy and the effect of timing of the surgery on physical and mental scores.

Methods

We analyzed data for 618 women from South Carolina enrolled in the ENDOW study. We assessed self-reported physical and mental health by using the Short-Form 12 (SF-12) and Mental Health Inventory 5 (MHI-5). Two sets of analysis of variance were conducted for each instrument with the following independent variables: hysterectomy status (had hysterectomy, doctor recommended hysterectomy, considered hysterectomy without a recommendation, and never considered), and hysterectomy timing. We controlled for education, ethnicity (black or white), and age.

Results

Twenty-seven percent of participants had had a hysterectomy, 60 percent had never considered it, and the other women were evenly divided between the remaining two groups. Among the women who had had a hysterectomy, 80 percent thought the timing was right, 13 percent thought it was too late, and 7 percent thought it was too early. Women who had never considered hysterectomy had the highest physical health scores; we noted little difference among the women in the other three groups. The women who had never considered hysterectomy also had the highest mental health scores, and the women who had considered hysterectomy without a recommendation had the lowest scores. Women who said their surgery was too early had lower scores than did the other women who had had hysterectomy.

Conclusions

Differences that were statistically significant were small and may not be clinically significant. Preoperative interventions should take into account individual cultural, and demographic factors and the potential impact on physical and emotional well-being of the surgery decision. Symptomatic indications for hysterectomy and the surgery itself may affect women’s emotional and physical well-being. Women who have surgery before they are ready may require additional support, counseling, or other services to achieve optimal well-being. Women who consider hysterectomy without a recommendation constitute a group requiring further study. Follow-up of women with symptomatic indications for hysterectomy should include assessment of functional status, quality of life, and mental health.

---

 

Women’s Preferences for Involvement in Medical Decisions

Author:
Janet Y. Groff, MD, PhD, MSPH

Background

In the qualitative phase of the ENDOW study, women expressed strong interest in participating in decisions about hysterectomy and hormone replacement therapy. We explored women’s preferences for involvement in decisions regarding hysterectomy and their health in general, and looked at possible correlations with race and ethnicity.

Methods

We analyzed data from the ENDOW survey participants from the University of Texas Health Science Center at Houston and the University of Alabama at Birmingham. Our dependent variables were the Decision Making (DM) and Information Seeking (IS) subscales of the Autonomy Preference Index, and a clinical vignette. For DM and IS, higher scores indicated stronger patient autonomy.

Results

1,760 women were included in the analysis, and the mean age was about 46 years. 48.4 percent were African American, 37.4 percent white, and 14.3 Hispanic. Most of the women had some level of education: 17.5 percent had less than a high school education, 28.2 percent had a high school education, and 14.3 percent had some college. More than half the women (55.2%) were married, almost half (47.9%) had household incomes less than $20,000 per year, and only 35.2 percent rated their health status as very good to excellent. The mean DM score was 61.0 (SD 18.0), and the mean IS score was 87.3 (SD 10.9). The mean vignette score was 48.9 (SD 18.5). Univariate analyses showed all independent variables were statistically significantly associated with all dependent variables, but the strength of the associations were small. In multivariate analyses, education was most strongly associated with the dependent variables. After controlling for other independent variables, the addition of race or ethnicity contributed small percentages of the variances (ranging from 1-3%).

Conclusions

Women expressed a strong desire for medical information and involvement in medical care. For medical decisions relating to fibroids and hysterectomy, women wanted an equal role with physicians. The magnitude of the effect of race and ethnicity on involvement with medical care was very small. Shared decision making for elective hysterectomy decisions should be promoted.