Study design and protocol for moving forward: a weight loss intervention trial for African-American breast cancer survivors

Demographic data include name, address, date of birth, marital status, number of children, education, occupational status, annual income, and insurance status.

Diagnosis and treatment history (e.g., chemotherapy, radiation, previous and current adjuvant endocrine therapies) are collected from the treating oncologist and will be used for descriptive purposes only.

Respondents review a list of sixteen health problems (i.e., hypertension/high blood pressure, diabetes, arthritis) and report if they have ever been told by a doctor that they have this particular condition.

This questionnaire asks respondents to rate on a five point scale (1 = never, 5 = very often) how often friends and family have done or said certain things related to the respondents’ efforts to change their dietary or exercise habits. Social support for eating habits is measured using two five-item subscales (encouragement and discouragement), each calculated separately for friends and family. Internal consistency coefficients range from 0.73 to 0.87. Social support for exercise is measured using one 10-item scale, also calculated separately for family and friends [41].

The Physical Activity and Nutrition Self-Efficacy scale is an 11-item instrument that assesses the participant’s level of confidence that she can complete particular activities that promote weight loss [42]. This scale has adequate reliability, internal consistency, and construct validity, as well as good predictive validity among African-Americans.

These items come from the Robert Woods Johnson Active Where study [43]. Respondents rate their level of agreement (from 1 = strongly agree to 4 = strongly disagree) to four statements related to access to physical activity resources, five statements related to healthy eating resources, and five statements about perceived neighborhood safety. All scales show good internal consistency, with Cronbach’s alphas from 0.78 to 0.94 [43].

The intervention was developed based on formative qualitative work with African-American breast cancer survivors, followed by a pilot of the guided program that led to further refinements [30, 62].

The Moving Forward intervention integrates concepts from Social Cognitive Theory (SCT) [63] and the Socio-Ecological Model (SEM) [64, 65] to promote behavior change. SCT suggests that behavior can be explained by the dynamic interaction between behavior, personal factors (e.g., self-efficacy), and the environment (e.g., social support [66‐70]. The intervention also incorporates tenets of the socio-ecological model (SEM) [64, 65], a model that goes beyond individual-level variables and emphasizes that support from the larger social context is needed for long-term behavior change [71]. Accordingly, SEM posits that weight status, diet, and physical activity are influenced by individual (e.g., beliefs, taste preferences), interpersonal (e.g., social support, traditions and role expectations), and community factors (e.g., access to resources that support health promotion) [72]. Interventions hoping to promote long-term behavior change must address these three levels of influence [73, 74]. Moving Forward accomplishes this by addressing: (1) Individual factors - acknowledging heavier body image ideals, identifying and addressing personal barriers to behavior change; (2) Interpersonal - the importance of food in the African-American culture and finding ways to integrate this value with healthful eating; providing low-fat versions of culturally traditional “soul food” recipes; acknowledging and addressing family roles and family resistance/support to change; providing information on integrating healthful lifestyle practices for the family; facilitating social support for making changes in diet, physical activity, and weight; understanding the important role of religion and worship in the women’s lives and how it affects their health perspectives and (3) Community - incorporating a sustainable link to a community physical activity resource that can address barriers to regular physical activity (i.e., safety, weather, access); problem solving around cost and availability of healthy food; introducing participants to unfamiliar community resources. Interestingly, a positive sense of community (e.g., social bonds between individuals and between individuals and their community) is associated with self-efficacy for physical activity among African-American women [75].

The overall goal of Moving Forward is to make changes in health behaviors to promote a healthy weight. The weight loss goal is 7 % of baseline body weight (1–2 lbs. per week), consistent with the recommendations of an expert panel at National Institutes of Health [76]. Dietary goals aimed at producing weight loss, decreasing breast cancer recurrence risk, and improving overall health include: (1) a decrease in daily caloric intake (based on weight in pounds X 12 caloriesl/day, with 500–750 cal subtracted to create an energy deficit); (2) a decrease in dietary fat consumption to 20 % of total calories; (3) an increase in fruit and vegetable consumption to 7 daily servings; and (4) an increase in fiber to 25 g per day. For exercise, participants are advised to gradually increase their activity to a minimum of 180 min per week at 55–65 % maximal heart rate.

The MFG program meets twice a week for 26 weeks (see Table 1 for weekly themes) and is led by a team that includes a community dietitian, a community cancer exercise instructor, and a health psychologist. The program is conducted in city park district facilities, where participants enjoy free memberships, ongoing access to classes and fitness rooms, and the opportunity to maintain contact with program participants once the program concludes. The first meeting each week includes a 60-min class that addresses knowledge (e.g., relationship between obesity and breast cancer; food label reading; portions; available healthy living community resources), attitudes (e.g., pros and cons of weight loss; understanding the roles that food plays in one’s life; the concept of “fail to plan, plan to fail”) and cognitive behavioral strategies including self-monitoring of weight, food and physical activity; realistic goal setting; stimulus control; problem solving; mindfulness; cognitive restructuring and relapse prevention. These classes are led by a dietitian, a psychologist and a certified cancer exercise trainer. Table 1 provides a list of weekly curriculum topics. Pilot data showed that many women entered the program with low levels of knowledge about healthy eating and exercise. Thus, the first weeks are devoted to teaching core concepts (e.g., concept of calories in/out, food label reading, measuring heart rate), while later weeks are focused on cognitive-behavioral aspects of weight loss such as stimulus control, habit and mindfulness. Class activities include weighing in weekly; completing a food and activity self-monitoring record; increasing awareness of portions by weighing and measuring foods according to one’s typical portions and then according to recommended portions; creating stimulus control plans for home, car and work; identifying barriers to healthy eating and/or exercise and problem solving within small groups; going on a field trip to a local grocery store to practice reading food labels; creating an eating-out management plan; and identifying high-risk situations and brainstorming ways to manage them. The first weekly meeting also includes a support “icebreaker” (share the funniest moment of your breast cancer journey; what has been the most frustrating; etc.) and a 60-min exercise class taught by a certified cancer exercise trainer.

The second weekly meeting is a 60-min exercise class taught by a cancer exercise trainer. The exercise classes incorporate a variety of activities, including traditional aerobics, line dancing, African dance, salsa, yoga, Pilates, and strength and flexibility training. Class time is also spent learning to use the park district fitness facility equipment to ensure that women feel comfortable and competent on the equipment, thus promoting enhanced self-efficacy and mastery of new skills. Many participants enter the program at very low levels of fitness; therefore, physical activity levels are increased gradually with special attention to concerns such as lymphedema and balance. Increased physical activity outside of class is encouraged by suggesting enrollment in additional local exercise classes, providing safe outdoor walking routes, and alerting women to activity resources online and on FitTV.

Participants often need further support and reinforcement of lifestyle changes outside of class, as well as timely information related to healthy eating and exercise resources. To provide this efficiently and effectively, MFG uses text messaging, a strategy successfully used in previous weight loss interventions with low income African-American women [77, 78]. A custom software application, mytapp, allows the psychologist and trainers to schedule two text messages each week to be sent to participants. Messages are less than 200 characters in length and are written to be brief, clear and motivational. The intent is to reinforce concepts covered in class while also supporting self-efficacy, social support and perceived access. For example, during the week that the intervention covers portions, participants receive a text that reads, “Are you using your measuring cups and spoons?” In the week that we cover mindfulness, the message states, “Breathe, Breathe deeply. You CAN savor each bite.” At the end of the six-month program, MFG participants receive six monthly newsletters to review and support concepts related to integrating and maintaining healthy lifestyle habits. They also receive supportive weekly text messages until the final data collection.

Participants randomized to the SG receive a program binder that is divided into 26 sections, with each section addressing the topics listed in Table 1. Within each section is a brief guide to the topic and accompanying worksheets, handouts or activities to complete. Administrative study staff (not intervention staff) meet briefly (approximately 15 min) with the SG participants once individually to give them their binder and to provide an orientation to its layout and contents. Staff follows up with a monthly check-in phone call. At the end of the six months, SG participants also receive the monthly newsletters for six months until the final follow-up data collection.

Calculations were conducted for hierarchical longitudinal designs with differential attrition rates using software provided by Roy et al. [79] Results showed that an initial sample of 240 participants is sufficient to detect an effect size of 0.4, assuming attrition of about 10 % at each time point with power of 0.8 for a two -tailed test. Power will increase with the inter-correlation of observations over time, which we estimated conservatively at 0.4. The addition of covariates to the model will also increase power.

For a two-group randomized trial with repeated measures the mixed (multilevel) model is the preferred mode of data analysis [80]. With only two time points, the analysis reduces to a test of the difference between treatment groups in gains or losses in the outcome variable between baseline and 6-month follow up. The model easily generalizes to three time points (including the 12-month visit), in which case time can be indexed in simple linear fashion or represented either in quadratic form or via indicator variables for the first and second follow up measures. In that case, the analysis focuses on between group differences in outcome trajectories over time.

Our hypothesis is that subjects in the MFG condition will experience greater rates of weight loss, improved diet and physical activity patterns and other positive outcomes than subjects in the SG group. A major advantage of casting data analysis in the framework of the generalized linear (regression) model is that it easily accommodates covariates. One important covariate is time since diagnosis, which will be included in all analyses. A potentially important aspect of our modeling will be the investigation of dose–response effects. We will have data on frequency and pattern of attendance for each subject, which will allow us to model outcomes as a function of intensity of treatment.