Some argue health disparities are merely a function of differences in socioeconomic status; however, there is growing evidence that structural racism contributes to racial health disparities (22). Structural racism includes policies and laws which provide advantages to groups based on race while oppressing another race (23, 24). Black adolescent females living in low-income communities are at particular risk for obesity due to racial residential segregation and its role in creating obesogenic environments characterized by reduced access to healthy foods, heavy marketing of unhealthy food (25, 26), few safe, walkable streets, park areas, and exercise facilities (13, 27–30). To prevent obesity and the associated co-morbidities in this at-risk population, programs must acknowledge the impact of racism on dietary and PA behaviors, rather than simply prescribing interventions that do not consider their lived racialized context.

Adolescent daughters often rely on their mothers for information, guidance, and support as they work to adopt or maintain healthy behaviors (31). Our cross-sectional, descriptive study to assess the eating behaviors of 10- to 12-year-old Black daughters and their mothers (N = 43) showed strong concordance between daughters’ and mothers’ consumption of discretionary calories, such as sugar sweetened beverages (32). Similarly, the influence of mothers in shaping their daughters’ dietary and PA behaviors has been demonstrated repeatedly in cross-sectional studies (6, 25, 33). We developed BGM by acknowledging the environmental and developmental factors that affect Black adolescent females and their mothers. The complex intersection of environmental and developmental factors has exacerbated barriers to the recruitment of Black adolescents into research studies. These recruitment barriers can be structural (transportation, inconvenience) or cultural (lack of cultural concordance of mistrust) and require innovative and tailored recruitment strategies (34). Similarly, Black women face recruitment barriers, such as inability to miss work or lack of trust in researchers (35).

A review of randomized controlled trials (RCTs) of culturally tailored obesity treatment interventions (N = 44 RCTs) for adolescents aged 12–17 years, found low quality evidence for reducing BMI, inconsistent evidence for improving dietary behaviors, and no evidence for improvements in objective or self-report measures of PA (36). Similar findings were reported in two relevant obesity treatment RCTs Go Girls (37) and the Health Improvement Program (HIP) for teens (38), that tested culturally tailored, family-focused interventions (primarily involving mothers) in overweight and obese Black adolescent females in the US. Go Girls (37) was not effective for reducing adiposity, while the HIP for Teens showed an initial reduction that was not maintained. Several factors may account for the limited effects of these two treatment interventions: (1) once obesity is established in childhood it is hard to reverse (39), which supports our intervention to prevent obesity in Black adolescent females; (2) because Black adolescent females may show greater acceptance of heavier bodies (40, 41), weight maintenance may be a more acceptable goal than weight loss; and (3) parents participated in both interventions, but there was no focus on modifying daughter-mother communication, problem solving, role assignment, or relationship quality.

Adolescence is a critical transitional period marked by growing autonomy, yet it remains a time when parental influence is still essential. It is important to establish relevant recruitment and retention plans for obesity prevention interventions with Black adolescent females. There are known barriers to recruitment of Black adolescents (42), Black mothers, and even Black youth/parent dyads (43). However, less is known about specific barriers to recruitment of Black adolescent daughter/mother dyads. Further, obesity prevention studies with Black daughter/mother dyads to date have not focused on the pivotal developmental period of adolescence. Given the lack of obesity prevention research with Black adolescent daughter and mother dyads, the purpose of this paper is to describe the lessons learned from recruiting participants into the first 2 years of Black Girls Move (BGM), an obesity prevention/weight maintenance intervention.

We designed Black Girls Move (BGM) loosely from the Diabetes Prevention Program, using community engaged research principles of working with communities of interest. We first used interviews and focus groups to draft and refine BGM with Black adolescent daughter/mother dyads and clinical experts (44). We then pilot tested BGM with Black adolescent daughter/mother dyads and incorporated their feedback into its refinement. This iterative process to intervention development produced a culturally tailored and informed intervention (45). Participants provided feedback on placement of content, delivery format, and appropriateness of materials (44). In addition, a culturally acceptable approach in BGM was the concept of weight maintenance, which has shown promise as an effective and culturally relevant strategy to address risk for obesity-related diseases in Black females (46). Based on participant feedback, we defined obesity prevention as encompassing both weight loss and weight maintenance.

BGM is a 12-week race-conscious, multicomponent, mobile health (mHealth) obesity prevention intervention designed for 7^th–10^th grade Black adolescent females and their mothers. Inclusion criteria for daughters were (a) English-speaking; (b) Black; and (c) daily access to the internet outside of school and/or work through an iOS or android smartphone, tablet, or personal computer. Inclusion criteria for mothers were (a) English-speaking; (b) Black; (c) co-residing biological mother or mother-figure and legal guardian of the participating daughter; (d) the person primarily responsible for meals in the household; and (e) access to the internet through an iOS or android smartphone, tablet or personal computer. Originally, we only included daughters that measured as overweight to focus on obesity prevention. However, this criterion excluded 21.6% of daughters (Figure 1). Therefore, we modified our study protocol to include all weight categories with medical release. In addition, the original protocol excluded 7^th and 8^th graders, which resulted in 47.4% of daughters being ineligible for the study. After speaking with key personnel in schools that served 7^th–12^th grades, we recognized the importance of establishing healthy behaviors in 7^th and 8^th graders and modified our study protocol to include 7^th–10^th graders. The initial recruitment results enabled us to refine inclusion and exclusion criteria to optimize participant safety, recruit sufficient numbers of participants, and promote generalizability of findings (47). In a future study we plan to include a longer follow-up period of 11^th and 12^th graders to assess maintenance of behaviors as adolescents transition to post-secondary activities. The original and modified inclusion criteria are further described in Table 1.

In addition to using mHealth technology, BGM actively engages Black adolescent daughters and their mothers through structured group sessions. The feasibility, preliminary change in outcomes, and impact of BGM will be evaluated against a daughter-only comparison condition (DOCC), which delivers identical content as BGM but excludes maternal participation. In the protocol, eligible participants are randomly assigned 1:1 to either the BGM active treatment or DOCC. BGM is informed by race-conscious, Public Health Critical Race Praxis (PHCRP), Family Systems, and Social Cognitive theories. The intervention development, description of each theoretical component, and outcomes of BGM are described in the BGM protocol (45). The specific aims are to determine in Black adolescent females:

(1) The feasibility (recruitment, retention, and acceptance) of BGM compared to DOCC.; (2) The preliminary change in primary outcomes, including: (a) number of steps per day, and (b) diet quality, measured by the self-reported intake of five components of the Healthy Eating Index-2010 (vegetables, fruit, whole grains, dairy, and protein), as well as secondary outcomes such as: (c) minutes of moderate/vigorous physical activity per week, (d) self-report physical activity, and (e) consumption of sugar sweetened beverages; and (3) The impact of BGM compared to DOCC on differences in theoretical mechanisms of change, including (racial identity, daughter/mother relationship, social cognitions) assessed by self-report measure.

As a feasibility study, we did not perform a formal power analysis to determine sample size. Instead, sample size estimates from the Diabetes Prevention Program were used, which suggested at least 5 participants per group to promote peer support, interaction, and group cohesion (49, 50). The group format with dyads provided a mechanism for mothers to share successes and adaptive response strategies with their daughters. Originally, if a minimum of N = 12 dyads per school/per condition was not met, the school was canceled but based on the challenges of meeting a total of 24 dyads per school we reduced the cancelation criterion to <N = 6 dyads. Six dyads were selected as a minimum recruitment target to account for absences, yet small enough to allow participants time to discuss new foods, ways of preparing foods, or new PA strategies with other dyads.

Prior to starting recruitment, we reviewed relevant literature to identify successful culturally tailored recruitment strategies with Black adolescents and determined that building trust, combined with active and passive recruitment strategies, was essential (51–53). Culturally tailored strategies to build trust include community engagement and partnerships, effective communication, culturally competent recruitment materials, involvement of role models, accessibility and convenience, ongoing support, and feedback and adaption (34, 54–56).

Building trust requires a thoughtful and respectful approach that integrates clear communication, cultural sensitivity, and meaningful community engagement. Toward that end, we were intentionally mindful of language nuances that could be ambiguous and lead to unintended connotations. We also cultivated trust by becoming familiar with the community—being present in local spaces, such as connecting with schools and consulting community leaders. We employed team members who shared cultural similarities with participants, and we trained them to interact respectfully with families and school personnel to forge and sustain positive relationships. We conducted in-person enrollment and approached informed consent with care and respect (34, 54–56). Finally, we held in-person meetings with key school leaders to assess the communication and language preferences and norms of students, teachers, and staff.

Active recruitment refers to direct interaction with the target population to increase awareness about the study and provide prospective participants the opportunity to approach the researcher(s) (53, 57, 58). Our active recruitment plan consisted of multiple components. First, we obtained endorsements from the school principals. Second, we built relationships through separate recruitment activities for daughters and mothers. Recruitment activities with daughters occurred during lunchroom or after school activities, while activities with mothers occurred during parent report card pick up. Third, we encouraged daughters to fill out a sheet that collected personal information of daughter (first and last name, grade, and phone number) and mother (first and last name, email, phone number, and preferred time of contact). This information was used to contact mothers to begin the screening process for eligibility. During encounters with both daughters and mothers, a research team member screened for eligibility, explained the study, and scheduled a baseline assessment if time permitted. Otherwise, screening and study explanation were completed on a subsequent telephone call. Lastly, in-person presentations were given to mothers at parent advisory board meetings and to daughters and mothers at a college interest fair. The active recruitment strategies allowed us to obtain contact information for subsequent telephone calls to screen for eligibility. In addition to these strategies, we were responsive to school-specific requests related to relationship building. For example, at two of the high schools, the research team held a health careers seminar during the “flexible period.” At another school, health careers information (59) was presented during the school’s parent night.

Passive recruitment refers to increasing awareness about the study without direct contact (53). Our passive recruitment techniques included distributing flyers through e-mail and posting them in the administrative office, library, and gymnasium of 4 high schools (grades 9–12) and one middle-high school (grades 7–12). The flyer included the Black Girls Move logo, Quick Response (QR) code, information about the study, eligibility criteria, amount of participant gift card compensation for their time, and the principal investigator’s contact information. In addition to trust building, active, and passive recruitment (57, 58) study staff completed training specific to recruitment and retention strategies for minoritized populations. These strategies included training on historical fears of participating in research, partnering with community organizations, and tailoring recruitment materials (60).

Daughter/mother dyads were recruited from five Chicago Public High Schools (one of the five included a middle school). Each high school had a student body of greater than 80% of Black students, poverty rates above 80%, and was not on academic probation. While these criteria limited the number of schools that were eligible to participate in the study, they assured that we recruited from the targeted population. The Consolidated Standards of Reporting Trials (CONSORT) (61), was used to document the flow of participants through the recruitment process. Recruitment data were maintained in a secure Health Insurance Portability and Accountability Act (HIPAA)-compliant database, created in Access for Microsoft 365. The database contained participants’ names, phone numbers, dates of birth, and daughters’ grades and schools. The data manager trained research assistants (RAs) to input data into the Access database using the RA training manual developed by the principal investigator. RAs used the Access database to document the outcome of calls, the number of call attempts, and enter eligibility criteria. RAs followed a documentation protocol for each recruitment activity. For example, lunchroom recruitment protocol included guidelines for RAs to enter participant data into the Access database within 48 h after lunchroom recruitment and to send text messages to all mothers according to the communication protocol. Once contact was made, RAs would describe the study, answer questions, and screen those interested according to the six phases of the communication protocol (Table 2).

The principal investigator reviewed the CONSORT diagram weekly. Results were shared with co-investigators in weekly meetings to monitor accrual against targets. Enrollment rates were calculated and reported to the funding agency annually. Assessment periods for each school lasted for 8 weeks. The first 4 weeks included lunchroom and after-school recruitment, followed by 4 weeks of baseline screening. Recruitment and enrollment rates for each school were evaluated at the end of each assessment period. Results of recruitment and enrollment were used to plan potential modifications to recruitment methods.

Two hundred and ninety-four daughters expressed interest and received telephone screening call attempts (Figure 1). Of the 294 dyads, 44.2% (n = 130) could not be reached for telephone screening. Of the 164 mothers screened, 59.1% (n = 97) were ineligible (i.e., not in 9^th or 10^th grade, or BMI > 95%) and 32.3% (n = 53) were excluded (i.e., missed assessment/enrollment period or no response). Fourteen dyads were consented and assessed for eligibility at baseline screening. Of the 14 dyads that were screened at the in-person baseline screening, 14.3% (n = 2) were ineligible for reasons that were not documented and the remaining 12 dyads were assigned to a condition. Of the 12 dyads, one failed to return provider consent and one failed to wear the initial actigraph. Our feasibility data showed that the recruitment goal of 24 dyads per school (12 dyads per condition) was not met, therefore we canceled the intervention at four schools. The last step of the guidelines for evaluating feasibility of recruitment involves modifying recruitment methods based on the tracking data. Our modifications are described below.

Cunningham-Erves et al. (60), applied a community engaged approach to develop the Community-Informed Recruitment Plan Template, hereafter referred to as, the template. The template serves to help increase recruitment and retention of racial and ethnic minority groups by engaging community members in targeted discussions. The template includes: (1) Recruitment Strategy; (2) A Stakeholder Communication Plan; (3) Evidence of Recruitment Feasibility; (4) Recruitment and Retention Team; (5) Recruitment and Retention Methods; (6) Recruitment and Retention Timeline; (7) Evaluation; and (8) Budget. We applied the template to develop a plan for study modifications.

Despite our recruitment efforts, we did not meet recruitment targets in Year 1. Hence, we analyzed tracking data, met with community leaders, and engaged an expert community research consultant to modify our recruitment methods. The consultation yielded four lessons that influenced recruitment changes: (1) the importance of maintaining relationships over time to sustain trusting relationships, (2) prior weight eligibility criteria were too prescriptive, (3) the investigators need to change recruitment targets to minimize school loss, and (4) competing demands influenced engagement. Consequently, we (1) hired a community health worker (CHW) to address communication and scheduling concerns and to maintain trusting relationships with eligible participants until the intervention started; (2) amended eligibility criteria to include daughters of all weight status and 7th & 8th graders; and (3) increased incentives to $40 per participant at each data collection point, from $20 per participant to align compensation with the time requested to engage in the baseline assessment process.