Support Needs of Overweight African American Women for Weight Loss

Abstract

Objectives

To examine social support needs of obese and over-weight African American women for weight loss.

Methods

Focus groups were conducted with over-weight and obese African American women. Data were analyzed using standard grounded theory text analysis.

Results

Our middle-aged (45.7 years; SD=12.6) women (N = 66) were interested in receiving support from others focused on the health benefits of weight loss. Behaviors perceived as supportive include co-participating in exercise, providing nutrition education, using positive reinforcements, and avoiding criticism. Conclusions: African American women are interested in a program designed to increase social support for their weight loss.

The prevalence of obesity has reached epidemic proportions.¹ Second only to tobacco as the leading cause of death, obesity is a preventable cause of morbidity, mortality, and excess health costs in the United States.² Although highly debated,^3-5 the annual number of deaths attributable to obesity among US adults is approximately 112,000.⁴ Compared to persons of healthy weight, over-weight individuals have an increased risk of premature death from all causes with the largest mortality risk being cardiovascular disease. Obesity has also been causally associated with type 2 diabetes, several cancers (eg, colon), osteoarthritis, obstructive sleep apnea, respiratory illness, poor female reproductive health, and stress incontinence.⁶ Medical expenses attributable to obesity reached an estimated $75 billion in 2003, which suggests that the total direct and indirect costs of obesity may now be as high as $129 billion dollars.⁷

The obesity epidemic has disproportionally affected minority and low-income populations.⁸ In particular, African American adults are more likely to be obese (30.4%) compared to European American adults (20.8%). The higher prevalence of obesity among African Americans is a contributing factor to higher total mortality and elevated rates of coronary heart disease, diabetes mellitus, and hypertension.⁹ This racial/ethnic disparity has been observed primarily among women, with 67.3% of African American women qualifying as overweight or obese, compared to 46.8% of European American women.

Prior research has indicated that obese African American women want to lose weight,^10,11 but less than half of those attempting to lose report being able to obtain professional help to do so.¹² Further, in controlled clinical trials, African American women have been less successful in their weight loss efforts when compared to European American women.¹³ The lack of success in traditional programs has been hypothesized to result from discrepant value assumptions underlying mainstream obesity treatments as compared to those of African American groups.¹⁴ That is to say, the dominant culture and traditional programs link attractiveness with thinness. However, African American women are more likely to experience cultural pressure to be self-accepting of their physical shape, regardless of size.¹⁵ Further, African American women tend to prioritize the needs of peer and extended family networks over independent, self-determined goal setting.¹⁶ Therefore, treatment programs that assume participants are motivated to lose weight in order to feel more attractive or use independent goal setting as the primary change strategy might not be aligned with the cultural perspectives of African Americans.

Few studies have empirically examined racial/ethnic differences in attitudes and beliefs about weight loss efforts. One recent study found that compared to rural European American women, urban African American women reported less social support and role-modeling, lower self-efficacy and outcome expectations, and use of fewer behavioral strategies for weight loss.¹⁷ Another study compared African American to European American women matched on age, education, and income and found that African American women experienced less social pressure to be thin.¹⁸ Given the documented importance of social support in weight loss efforts, limited social reinforcement for thinness may undermine weight loss efforts among African American women.

Involving support partners in behavioral weight control treatment is one of few strategies shown to improve weight loss maintenance.^19-21 However, in most studies examining the impact of support partners in weight loss treatment conducted to date, approximately 95% of participants were white. Recently, Gorin and colleagues²² reported that participants (approximately 85% white with at least one successful weight-loss partner (ie, weight loss ≥ 10% at 6 months) lost significantly more weight at 6, 12, and 18 months than did those without successful partners and those without partners. Although bringing support partners to treatment represents a potentially promising way to improve both initial weight loss and maintenance of weight loss, there have been exceptions in the literature.^23,24 A meta-analysis of 13 studies examining the impact of spouse involvement in weight-loss treatment reported only a modest, short-term effect of spouse involvement.¹⁹ Perhaps this is because as a group, spouses are not always interested in helping their partners lose weight.^20,25 Hart and colleagues²⁶ suggested that uninterested spouses and significant others actually have a negative impact on weight loss efforts.

To date, there have been few studies that have compared the effectiveness of natural social support between participants and their family or friends with that of experimentally created social support, where strangers are encouraged to support one another.²² Further, no identified studies have examined the possibility of differential impact of social support on weight loss among African American samples. Given the value of the family and extended network in the African American community, targeting social support for weight loss may be an innovative method to address disparities in weight loss among African American women.

To inform future intervention work, the aim of the present study was to further our understanding of the knowledge, attitudes, and beliefs of overweight and obese African American women regarding support needed during a weight loss effort. Specifically, we sought to understand preferences for supportive behaviors for weight loss and physical activity and the potential role of a “supportive other” to assist in weight loss efforts.

METHOD

Participants

African American (AA) women were recruited and focus groups conducted at a community-based medical center serving urban, underinsured, low-income patients in Kansas City, Missouri. Patients were recruited via flyers posted in the adult medicine clinic, pharmacy, and lobby of the medical center. In addition, clinic staff were informed about the study's eligibility criteria and enrollment process so that potential participants could be referred by their primary health care provider. The one-page recruitment flyer was designed to recruit AA women who “thought they might be overweight” and “were interested in participating in a discussion about weight and weight loss.” Participants were required to be female, at least 18 years old, consider themselves to be black or AA, have a valid address and phone number, have a body mass index >25 (to be eligible to participate in an “overweight” focus group) or ≥30 (“obese” focus group), provide the name and location of their primary physician, and be available to attend a group discussion pertaining to weight loss. Study exclusion criteria included living in a shelter, halfway house, or detox center, or having a serious medical condition that might interfere with completing the focus group discussion. All procedures and protocols were approved by the Human Subjects Committee at the University of Kansas Medical Center.

Of the 97 individuals who responded to the study flyers, 96 were eligible. One potential participant was excluded given a BMI<25. Of the 96 eligible participants, 30 did not show for their scheduled appointment. The overall focus-group attendance rate was 68.7%. Specifically, 92.3% of those invited to attend an obese focus group attended (48/52), and 75% of those invited to an overweight group attended (18/24). Thus, a total of 66 African American women attended a focus group discussion. No significant differences were noted between those who attended and those who did not in demographic information gathered during screening.

Procedures

Prior to the onset of the focus groups, the screening instrument, survey, consent form, and moderator's guide were pilot tested and revised based upon the feedback from our study volunteer, an obese, African American female representative of the health center community.

Women who responded to the flyer were screened for eligibility by phone. Eligible participants provided contact information and available times and were contacted by phone within 2 weeks to be scheduled for a group. Focus groups were stratified by BMI into 2 groups (ie, obese women [BMI greater than 30] and overweight women [BMI >25 and <30]). Although women were invited to attend a group depending on their verbal report of height and weight (corresponding BMI calculated), participant actual BMI was determined by a body composition scanner prior to study enrollment.

Six focus groups were conducted in the obese stratum (n = 48) and 4 in the over-weight stratum (n = 18) over a 5-week enrollment period. One clinical psychologist with training in group facilitation moderated the focus groups. In addition, 3 research assistants with training in the study protocol, focus group facilitation, and protection of human subjects helped by welcoming participants, offering them healthy snacks, completing forms, video-and audio-recording, and distributing incentives. One research assistant was an AA woman from the local community who was employed by the research team for more than one year and whose primary task was to assist with focus group studies. Prior to the onset of the session, a research assistant measured each participant's height and weight behind a privacy partition to corroborate her self-reported data and to ensure that each participant was assigned to the correct group (ie, obese or overweight).

Prior to data collection, all participants provided informed consent and permission to be audio-and video-recorded. Our discussions took place around a large rectangular table, and participants were given the option of sitting wherever they chose. Prior to the onset of the discussion, participants completed a 15- to 20-minute survey documenting demographic and weight-related information. One research assistant read all questions aloud while another circulated to assist individual participants as needed.

The focus groups followed a semistructured format with a standard guide of open-ended questions to stimulate discussion about support for weight loss. Researchers with experience in weight loss and social support developed the guide according to published methodological suggestions.²⁷ Table 1 displays questions from the focus group guide. During each focus group meeting, the moderator probed participants’ responses and encouraged all members to participate. One of the research assistants took detailed notes, provided a summary at the end of the meeting, and encouraged participant reflections on the accuracy of the summary. The group discussions lasted approximately 90 minutes and were audio- and video-recorded. At the conclusion of the discussion, participants received a $30 Wal-Mart gift card as compensation for their travel cost, time, and effort. We stopped data collection after we felt that data saturation had occurred for the majority of our topic areas (ie, no new data would be found by conducting further focus groups²⁸).

Table 1.

Questions from Moderators Guide

Topics	Questions
A. Opinion of the term “obese” and preferred terms	1.Tell us your thoughts about the terms people use for weight. 2.What does the word ‘obese’ mean?
B. Community acceptance of overweight	1.Tell us about the people in your life. We are interested in whether you consider them to be overweight 2.How is being overweight viewed in your community?
C. Opinions about receiving help for weight loss as well as preferences	1.Have you ever asked anyone for help in losing weight? for types of help 2.Has anyone ever offered to help you lose weight? 3.If someone was concerned about you weight, what could they do to help you and what should they not do?
D. Opinions about providing help to another for weight loss ie, are they concerned about the weight of another?	1.Do you have any concerns about someone else's weight? 2.Have you ever discussed their weight with them, how did it go?
E. Preferred characteristics of support person	1. We understand that some relationships may be affected by one person helping another with changing behaviors associated with eating and physical activity. Which relationships would be positive and negative?
F. Cultural factors impacting someone supporting another's weight loss effort	1. Is there anything about the African American community that might be important in whether one person helps another lose weight?
G. Home rules about eating	1.What rules if any, do you have about food in your home? 2.How did these rules come about and how are they going?
H. Interest in program that would enroll support person and what should it include?	1.Would you be willing to participate in a training program that would be designed to teach another person how to best support you in a weight loss attempt? Why or why not? 2.What should be included in the program? 3.Would you prefer that this person was also trying to lose?
I. Interest in buddy system; who is the perfect buddy?	1.Would you like to have a weight loss “buddy”? 2.What would you like this person to do with or for you? 3.Describe the perfect weight loss buddy.

Measures

Quantitative survey

For purposes of this study, we developed a 20-item paper-and-pencil survey to document participant demographics including age, education level, marital status, monthly income, health insurance, self-reported health status, and comorbid medical illnesses (eg, diabetes, heart disease, cancer). Several psychosocial variables related to our study aims were also assessed. These included motivation and confidence to lose weight (0-10 scale, 10 = highest), number of lifetime weight-loss attempts, current weight-loss efforts, amount of support for weight loss received from family and friends (ie, none, a little/ some, and a lot), and most supportive relationship. (See Table 2 for items collected.) No attempts were made to validate the survey.

Table 2.

Demographics Characteristics of the Total Sample (N=66) and Among Those Who Participated in the Obese and Overweight Groups

Variable	Total N = 66	Obese n = 48	Overweight n = 18
Age, mean (SD)	45.7 (SD=12.6)	44.4 (SD=12.8)	49.4 (SD=11.5)
BMI, mean (SD)	36.9 (SD=9.2)	40.1 (SD=8.6)	28.3 (SD=3.6)
Education, # (%)
Some HS, no diploma	13 (20%)	7 (15%)	6 (33%)
Diploma/GED	19 (29%)	14 (29%)	5 (28%)
Post HS, Voc/tech, some college	26 (39%)	22 (46%)	4 (22%)
College degree	8 (12%)	5 (10%)	3 (17%)
Marital Status, # (%)
Married/cohabitating	17 (26%)	11 (23%)	6 (33%)
Divorced/separated/widow	23 (35%)	17 (35%)	6 (33%)
Single/never married	21 (32%)	17 (35%)	4 (22%)
Income, # (%)
< $1800 monthly household	43 (65%)	30 (63%)	13 (72%)
Health Insurance, without, # (%)	17 (26%)	13 (27%)	4 (22%)
General health, # (%)
Excellent/Very Good	20 (30%)	11 (23%)	9 (50%)
Good/Fair	43 (65%)	35 (73%)	8 (45%)
Poor	3 (5%)	2 (4%)	1 (6%)
Co-morbidities, # (%)
Hypertension	32 (49%)	23 (48%)	9 (50%)
High Cholesterol	22 (33%)	16 (33%)	6 (33%)
Asthma/Emphysema	19 (29%)	14 (29%)	5 (28%)
Heart attack/disease	4 (6%)	1 (2%)	3 (17%)
Cancer	8 (12%)	5 (10%)	3 (17%)
Diabetes	15 (23%)	8 (17%)	7 (39%)

Moderator's guide

Group discussions followed our semistructured guide as a general template with the use of probes to clarify and to add depth. Using the grounded theory approach, we built upon the moderator's guide following each group discussion. Our general topics/domains focused on the following: participants’ experiences with weight loss attempts, preferred methods for weight loss, perceptions of supportive and nonsupportive behaviors others might engage in during their weight loss efforts, attitudes about training a supportive other to assist them in a weight loss effort, barriers to including someone else in their weight loss efforts, and interest in participating in a weight loss intervention with a “buddy.”

Data Analysis

Data gathered in this study represents a triangulation of our multiple sources of information including audio- and videotapes of sessions, survey data, observational and summary notes, and member reflection of accuracy of summary notes.

Survey

Quantitative survey data was double-data entered and analyzed using SPSS Version 14.0.²⁹ Nonparametric statistics were used to describe participant characteristics, and t-test and chi-square were used to determine if those who attended the groups differed from those who were invited but who did not attend. T-test and chi-square analyses were also used to compare obese and overweight participants on our descriptive measures.

Focus group transcripts

Audio-recordings of the focus groups were transcribed verbatim by a contracted professional transcription service. The focus group moderator proofread the transcripts and compared them to the video-recordings to check for completeness and accuracy. Analyses were led by a medical anthropologist with over 12 years of experience using qualitative methodology. An initial code book was developed based on the focus group moderator's guide, with codes for all major topic areas. Transcripts were separated by stratum and assigned to 3 independent coders who were instructed to deductively categorize using the initial code book. Simultaneously, they inductively coded by hand within each major topic area using a grounded theory approach whereby categories and concepts emerge from the text and are then linked.²⁸ The lead analyst then cross-checked all codes to identify minor discrepancies in the coding and terminology.³⁰ Few discrepancies were found in major topic (domains) and sub-topic areas, and those identified were discussed by the research team until consensus was reached. The lead analyst then developed preliminary thematic statements for saturated themes in each stratum based on the coding.³¹ Finally, the research team met as a group to discuss the thematic statements and refine their wording; any additional discrepancies were few and were resolved through consensus.

RESULTS

Participants

As detailed in Table 2, a total of 66 women participated in our 10 total focus groups (ie, 6 obese focus groups [n = 48] and 4 overweight groups [n = 18]). The average age of the women in our sample was 45.7 years (SD=12.6), and the average BMI was 36.9 (SD = 9.2, Range 25-66.5). Half of the participants had at least some college education (51%); 26% were cohabitating or married; 35% were widowed, separated, or divorced; and 32% were never married. Approximately one quarter of the participants had no health insurance (26%). Most of the participants rated their health status as good or fair (65%), as opposed to excellent/very good (30%) or poor (5%). Common comorbidities included hypertension, elevated cholesterol, asthma, emphysema, heart disease or prior heart attack, cancer, and diabetes. Although current interest in weight loss was not an inclusion criterion, approximately 86% of participants were “trying to lose weight.”

As expected, BMI was statistically higher among those who attended the obese vs overweight groups (39.36 vs 28.75, F 10.79, P=.002). Further, obese partici pants had significantly higher ratings of motivation to lose (F 4.625, P=.035, M= 7.0 vs 6.4, respectively) although most participants were moderately motivated to lose weight (M = 6.8, SD 2.9) and confident in their ability to do so (M=7.7, SD=2.6). No other statistical significant demographic differences were found between the overweight and obese participants in demographic characteristics (P<.05).

Approximately 83% (n=15) of the over-weight and 88% (n=42) of the obese participants reported that they were actively attempting to lose weight. Approximately one quarter of obese and overweight participants reported that they were likely to receive “a lot” of support from friends during a weight loss attempt (28% and 25%, respectively), but fewer obese women reported the likelihood of receiving “a lot” of support from their family (29% and 39%, respectively). Participants were more likely to receive support from a child (33%), friend (17%), or husband or partner (15%) than a sibling (8%) or parent (11%).

Major Themes

As detailed in Table 4, 7 saturated themes emerged across both categories of focus groups.

Table 4.

Saturated Themes Across Both Obese and Overweight AA Women

Topics	Questions
A. Opinion of the term “obese” and preferred terms	1.The term “obese” has an extremely negative connotation. “Obese” individuals are thought to be those who are unhealthy, sloppy, and those who cannot function due to their excess weight. 2.Participants preferred use of terms such as “thick,” “big-boned” or “full-figured.”
B. Community acceptance of overweight	1.The media portrays the ideal woman as too thin. African American women do not want to be as thin as this ideal. 2.People within the African American community are accepting of others who are overweight.
C. Opinions about receiving help for weight loss	1.Participants want and feel they need support for weight loss, but want communication to be framed in a discussion focused on improving health, not focused on improving appearance. 2.Confrontational discussions are not helpful, but may be more likely to occur when talking to a family member.
D. Preferred characteristics of support person	1. A “supportive person” must be willing to work together with the overweight individual; have a close relationship with them and ideally, has successfully lost weight in the past.
E. Cultural factors impacting someone supporting another's weight loss effort	1.Church affiliation is important to African American culture and can serve as a support system. 2.The church often fails to support weight loss efforts because food is seen as an important part of fellowship and congregations often have high calorie meals together.
F. Home rules about eating	1.Few participants had home rules about what, when, or where someone can eat. 2.Rules were more about home cleanliness than losing weight.
G. Interest in program that would enroll support person and what should it include?	1.A program for weight loss support could include the support person alone but would ideally include a pair of “buddies.” 2.The program would include comprehensive information about exercise, nutrition, emotional eating, and other aspects of weight loss.
H. Interest in buddy system; who is the perfect buddy?	1. They would enroll together and be paired to go through the process together.

Opinion of the term obese and preferred terms

Overall, participants believed that the terms obese and obesity were unacceptable and did not apply to them. Both overweight and obese participants reported that these terms have an extremely negative connotation. They indicated that the term best describes an overweight person who is “unhealthy, sloppy, and unable to complete activities” of daily living. Furthermore, individuals who are labeled “obese” are considered to be “diseased, disabled, and out of control.” Participants in both groups reported that they prefer terms such as “thick,” “big-boned,” or “full-figured” to describe themselves. Whether or not an individual is physically limited due to their weight was cited as the standard criteria to determine obesity.

Fat is fat or overweight. It's not obese because if that was the case we wouldn't be able to walk or talk.

(Obese, 24 y.o.)

Thick is a better word than fat.

(Obese, 71 y.o.)

Community acceptance of overweight individuals

Participants expressed a sense of acceptance of larger body sizes, both in terms of attractiveness and perceived health. They stated that attractiveness does not depend on being a particular size. Although participants expressed tolerance of larger body sizes, most acknowledged that the media portray the ideal woman as having a BMI less than 25 (“sickly thin”). Participants expressed the view that being healthy does not depend on being a certain size and that a person can be just as unhealthy if she is underweight. In fact, most expressed the belief that those who were underweight were more vulnerable to illness.

Everybody is accepting of it, so it's okay.

(Obese, 37 y.o.)

I feel like we're okay but according to the chart at the doctor's office, we're all overweight.

(Overweight, 22 y.o.)

Opinions about receiving help for weight loss

Participants acknowledged that they both want and need support for weight loss. However, many reported a lack of support for weight loss among their family and friends (eg, doubt about ability to lose, disapproval of weight loss efforts). Participants spoke about wanting supportive communications framed in a discussion of health improvement rather than enhanced appearance and preferring tangible forms of support (eg, working out together). Participants universally agreed that confrontational discussions were not helpful but were more likely to occur when talking about weight loss with a family member.

To me it's not a concern of being skinny; it's more about being healthy.

(Obese, 24 y.o.)

I need something and somebody that's going to be consistent and constant and help me maintain a healthy lifestyle.

(Overweight, 40 y.o.)

Preferred characteristics of support person

Participants expressed a preference for the following characteristics in a support person for weight loss: being empathic; being committed to working together with them to increase physical activity; having a close, caring relationship with them; using positive reinforcements and avoiding criticism; and having achieved successful past weight loss or other health behavior change themselves.

It's always got to be somebody that's been through the same problems that you're going through in order to help.

(Obese, 57 y.o.)

I think the best way they could support me is to have a relationship with me and to care about me.

(Obese, 44 y.o.)

Cultural factors impacting weight loss effort and support

Participants frequently mentioned their church or religious affiliations as a potential source of social support. However, they reported very mixed experiences regarding receipt of support from their church given the observation that high-calorie, high-fat “home-cooked” foods are often shared during fellowship meetings. Participants described social pressure to consume these items so as not to offend.

When it comes to church time you supposed to be able to eat what you want to eat. You might have dieted all during the week but this is church. It's time to eat.

(Obese, 45 y.o.)

Home rules about eating (stimulus control)

When asked about rules regarding the places or times when eating was sanctioned in their homes, few participants endorsed the presence of any home rules. When a rule about eating was in place in the home, it was typically designed to keep the home clean rather than as a method to control food intake (ie, stimulus control). Interestingly, participants universally endorsed the belief that limiting high-calorie (okay) foods from the home was not possible or appropriate if children resided in the home.

I don't have any rules. You know it's just a get in where you fit in.

(Overweight, 37 y.o.)

We don't have any eating zones or any of that kind of thing.

(Overweight, 31 y.o.)

Components of a support program

Participants expressed a strong interest in a weight loss program that would include enrolling as a pair to go through the process together (eg, “buddies”). They spoke about an interest in a structured, comprehensive program that would offer information about exercise and provide nutrition education, information on emotional eating and food preparation, and ideas to keep them motivated. Their preference was for a long-term (ie, 1-year) program with ongoing support from their buddy, both during the structured program and after.

The buddy system sounds good because you're there to support each other, to help each other.

(Obese, 36 y.o.)

I think education to let you know what you need to do, how you need to do it.

(Obese, 36 y.o.)

Differences between overweight and obese women

As detailed in Table 5, there were distinct differences in the focus group discussions between the over-weight and obese participants. Overweight participants described feeling that their weight loss efforts are not be taken seriously because they were not “really heavy.” Compared to obese participants, over-weight women were more likely to talk with male rather than female friends and to report that their spouses/partners were more supportive of their weight loss efforts. Finally, overweight women were more interested in concrete and specific advice for weight loss.

Table 5.

Theme Differences Across Obese and Overweight Participants

1. Overweight women were discouraged from losing weight by the majority of their obese friends and family member and told, “You don't need to lose weight.” As a result, they were concerned about mentioning their weight and weight loss efforts to others.

2. Overweight women were more likely to talk with male friends and family members than obese women.

3. Overweight women generally felt that they could do more to lose weight than obese women and were subsequently more interested in concrete, specific advice for weight loss, rather than general information.

DISCUSSION

To our knowledge, this is the first systematic investigation of the support needs of African American women for weight loss. Our qualitative and quantitative investigation revealed that both overweight and obese African American women were interested in receiving support in their weight loss efforts. This is consistent with results from prior qualitative studies conducted with overweight and obese samples in which the need for unspecified social support to assist in the weight loss process has emerged as a frequent theme.^15,32-34 For example, both Barnes and colleagues³² and Befort et al³⁴ conducted focus groups with African American women. The researchers asked group participants to describe key ideas that were relevant for weight loss. Individuals participating in these studies cited social support as being critical for maintenance of weight loss. The identification of social support as important for weight loss in these studies is further supported by past research that suggests that, in comparison to European American culture, African American culture subscribes to a more collectivist ideal that places emphasis on social groups as the central unit of concern.³⁵ A collectivist orientation translates into a focus on cooperation, responsibility for others, loyalty, helpfulness, forgiveness, family security, and respect for tradition,³⁶ with individual goals secondary to group goals. As described by Chatters and colleagues,³⁷ African American communities place higher value on creating supportive environments through extended family networks. Taken together with previous qualitative findings, social support may be a particularly important component to target when designing weight loss interventions for African American women. Family and friends may represent important agents of change to influence weight-loss health behavior change, but it appears that this resource is not currently being exercised. In our exploratory qualitative design, we attempted to further operationalize what is meant by social support to inform future intervention trails designed to target support as a mechanism to improve weight loss success.

Results from the present study revealed it is important that the individual chosen for support has an established, close relationship with the individual interested in losing weight. Additionally, characteristics of empathy, commitment, and prior weight-loss success were highly desirable. These results are consistent with those of Gorin and colleagues,²² who found that participants assigned to an exercise-intensive program who recruited at least one support partner with previous success at losing weight were more likely to lose weight themselves than were participants who recruited partners who were subsequently unsuccessful in their weight loss efforts. Thus, it appears that when a buddy system is instituted, the success or failure of the supportive partner has a direct impact on the participant. To our knowledge, no prior studies have explored support provided by a person not currently attempting weight loss.

Typically, African American women are not as successful as European American women at losing weight in traditional weight-loss programs.³⁸ Perhaps this is because African American women as a whole tend to be more accepting of their own body types and, as a result, have more positive body images than do European American women.³⁹ Consequently, a favorable body image may deter motivation to lose weight.⁴⁰ Participants in the present study indicated interest in losing weight as a means for improving their health rather than becoming thin. They also indicated the desire for support persons who would help them to improve their health and better maintain a healthy lifestyle.

Important behaviors believed to support weight loss included participating in physical activities together, providing nutrition education, using positive reinforcement, and avoiding criticism. The finding that overweight and obese participants desire positive support strategies is not surprising and parallels research examining the effects of social support on smoking abstinence. Cohen and Lichtenstein⁴¹ found data to suggest that enhancing positive and reducing negative behaviors of a support partner may promote and maintain smoking abstinence. In particular, they concluded that it was the ratio of positive to negative behaviors that better predicted abstinence than either the frequency of positive behaviors or negative behaviors alone. Drawing on this work, it may be especially important to teach support partners not only to increase behaviors seen as constructive but also to simultaneously decrease those behaviors seen as ineffective.

Participants acknowledged the role of the church in their lives and the potential for it to serve as a support system. Many previous studies have documented the positive impact of religious affiliation as an extended family network and source of informal social support for members of African American communities.^42-44 Given the importance of the church as a social institution in African American communities, it is not surprising that church settings have frequently been utilized to deliver prevention and intervention services including weight loss and exercise programs to African Americans.^33,45 However, participants in our study suggest that the social atmosphere surrounding church activities might also act as a barrier to weight loss. Specifically, food was described as an important part of fellowship gatherings that create solidarity among church members. These gatherings typically include potluck meals, to which church members bring high-calorie, traditional foods. Thus, although the church and its members may serve as a source of support for African American women in a weight loss attempt, this may not always be the case, particularly if church gatherings involve unhealthy food choices. Weight loss interventions developed for African American women would benefit from problem-solving methods to manage caloric intake during church functions. Further, it would be helpful to target church leaders to support inclusion of lower calorie dishes.

In an effort to understand how the family support environment may impact weight loss efforts, participants in the current study were asked to discuss home restriction regarding dietary intake. Our sample reported very few rules surrounding eating behaviors in the home. Some participants mentioned instilling rules in an effort to keep their homes clean, but participants did not indicate restrictions on when and where food could be consumed. These rules and strategies are positively related to increased weight loss.^46,47 For example, individuals exposed to behavioral weight-loss treatments are often instructed to implement specific rules such as eating only in one designated place and not eating in front of the television.⁴⁸ Although lack of eating restrictions and stimulus control procedures may not be specific to overweight and obese African American women, intervention approaches should consider the inclusion of behavior modification components in obesity prevention and weight loss interventions for African American families. These efforts would be substantially enhanced by including strategies to increase the observation of these home rules by family members.

A number of notable differences emerged when comparing our overweight and obese focus group themes. Overweight women reported a preference for support that included more specific and concrete advice for weight loss. The preference for more specific suggestions may be a reflection of the increased self-efficacy of overweight women that weight loss could be achieved. Interestingly, overweight women expressed a concern that because they were not as overweight as others, their interest in weight loss would be dismissed by others. In fact, overweight women mentioned that weight loss was discouraged by many of their family members and friends. Cultural and social norms surrounding weight in African American communities may be one possible explanation for this finding. Kumanyika¹⁰ suggested that African American women are surrounded by more tolerant attitudes towards obesity, which might contribute to the perception that one's efforts to lose weight will not be taken seriously. In fact, previous research has suggested that, in comparison to their European American counterparts, African American men prefer women who are slightly larger⁴⁹ and overweight. Further, African American women reported receiving more attention from men when they are heavier.⁵⁰ These findings suggest the need to train support persons to help negate the lack of perceived support that overweight women experience in their natural environment. The information generated from this and other qualitative studies can be used to inform targeted interventions that are sensitive to the specific challenges faced by AA women.

It is hoped that the themes that emerged in these focus groups will stimulate hypotheses for additional research and will inform the development of weight loss materials or intervention strategies designed to target support for weight loss among African American women. Because the participants were not a random sample of the population of overweight and obese African American women, the extent to which the findings can be generalized is limited. For example, attitudes and beliefs may differ among those willing to participate in focus groups versus those who are not able or willing to participate. Our results may apply only to women of lower-income status as all participants were recruited from a medical facility serving lower-income patients with few insurance resources. Further, our study design was restricted by unequal numbers of overweight and obese focus groups (ie, 4 vs 6 groups) and participants (18 vs 48 women). Our decision to include separate groups of overweight (BMI >25 to <30) and obese participants came after our first set of focus groups when we observed differing beliefs among those with BMI closer to 30 vs those with higher BMI. Although we intended to recruit similar numbers of participants, recruiting women in our restricted BMI range proved challenging. Although we completed recruitment of our obese participants after only 2 weeks of recruitment, we were able to enroll only 24 overweight participants after 4 weeks of recruitment efforts. In order to retain participants, we scheduled our “overweight groups” once enrollment reached 6 participants, as compared to 12 for our “obese groups”. Although differing numbers of people in each focus group have the potential to affect the response of participants, in qualitative research, the reliability of the study results are not dependent on the size of the sample but on achieving saturation or redundancy in the information gleaned.⁵¹ We were able to reach saturation after completing 4 overweight groups, whereas new information continued to emerge from our obese groups until the sixth group. It is unclear from our data whether our population of overweight women was simply limited in number or if women with BMI < 30 were less likely to see themselves as overweight and respond to our flyer. Nonetheless, given the paucity of research investigating the social support needs of African American women in a weight loss attempt, even with the relatively small sample size, this qualitative study adds to the literature regarding social support and weight loss among African American women and provides an initial formative step toward developing interventions.

Study results should also be interpreted within the context of the strengths and limitations of focus groups. Although focus group research is, by design, a qualitative approach that provides a more in-depth picture of the issues of interest and can be essential for informing the design of successful interventions, it is not designed to be generalizable. Because we reached saturation in our major themes, we believe the results to be transferable to similar populations and able to serve as a starting point for further research among overweight and obese African American women residing in similar areas.

Despite these limitations, the current study offers insight into the support needs of both overweight and obese African American women who are interested in losing weight and adds to the body of literature examining weight loss perceptions and beliefs among AA women. We plan to design and conduct an efficacy trial to determine whether an intervention designed to increase social support for weight loss is feasible, acceptable and results in increased weight-loss efforts and maintenance of weight lost.

Table 3.

Weight-related Characteristics of the Total Sample and Among Those Who Participated in the Obese and Overweight Groups

Variable	Total N = 66	Obese n = 48	Overweight n = 18
*Motivation to lose weight, mean (SD)	6.8 (SD=2.9)	7.0 (SD=2.7)	6.4 (SD=3.5)
Confidence to lose weight, mean (SD)	7.7 (SD=2.6)	7.4 (SD=2.7)	8.2 (SD=2.5)
Described self as “overweight”, # (%)	60 (91%)	44 (92%)	16 (89%)
# Prior lifetime weight loss attempts, # (%)	5.0 (SD=12.9)	5.3 (SD=14.4)	4.1 (SD=7.3)
Trying to lose currently, # (%)	57 (86%)	42 (88%)	15 (83%)
Most Supportive Person, # (%)
Child	22 (33%)	18 (38%)	4 (22%)
Friend	11 (17%)	8 (17%)	3 (17%)
Husband/partner	10 (15%)	5 (10%)	5 (28%)
Sibling	5 (8%)	5 (10%)	0 (0%)
Parent	7 (11%)	6 (13%)	1 (6%)
Other	5 (8%)	4 (8%)	1 (6%)
Support from family, # (%)
None	11 (17%)	7 (15%)	4 (22%)
A little/Some	34 (51%)	27 (57%)	7 (39%)
A lot	21 (32%)	14 (29%)	7 (39%)
Support from friends, # (%)
None	20 (30%)	13 (27%)	7 (39%)
A little/Some	29 (44%)	23 (48%)	6 (34%)
A lot	17 (26%)	12 (25%)	5 (28%)

^*F 4.625, P = .035