In total 772 records were identified from five databases, Google, and Google Scholar searches. After removing duplicates and conducting the screening of titles, abstracts, and full texts, 25 studies were identified for inclusion in this systematic review. The reasons for excluding studies were: (1) not conducted in OECD countries, (2) not peer-reviewed, (3) not an RCT, and (4) full-text not available. Search results at each stage of the review process are illustrated in Fig. 1, using the PRISMA diagram.

Fig. 1

Systematic search results in PRISMA flow chart.

Characteristics of included studies

This review included 25 studies on recruitment and retention of ethnic minorities in obesity prevention RCTs. Of the 25 studies, majority of the studies were conducted in the USA (N = 23) [2, 3, 8,9,10, 13, 30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45] except 1 that was conducted in UK [46] and 1 in Australia [47]. Seven RCTs were exclusively conducted among Latino Americans [30, 34, 38,39,40,41, 44], 4 among African Americans [31, 35, 37, 45], and 1 among Mexican descent immigrants in the USA [43] and 1 among Brazilian immigrants in the USA [11]; 3 RCTs combined Latino and African American population [8, 33, 36], 1 combined Hispanic and Latino population [32] and 1 combined Hispanic and non-Hispanic black population [3]. The remaining studies were a combination of populations from different ethnic backgrounds [2, 9, 10, 13, 42, 46]. The studies used various statistical techniques for data analysis, including descriptive statistics (proportion, percentage, mean, interquartile range, standard deviation), inferential statistics (Chi-square, Fisher’s exact tests, t-test, ANOVA, Wilcoxon and Kruskal–Wallis tests, regression analysis). Few studies did not mention any analytical techniques.

Recruitment and retention of ethnic minorities in obesity prevention RCTs

This review found that the pooled proportion of recruitment of ethnic minorities in obesity prevention programmes and trials was 64% (Fig. 2). Of those recruited, the pooled proportion of retention of ethnic minorities in obesity prevention programmes and trials was 71% (Fig. 3).

Fig. 2

Pooled analysis of the proportion of the sample that is ethnic minorities and migrants.

Fig. 3

Proportion of ethnic minority and migrants who completed the programme.

Barriers to effective recruitment and retention

The meta-ethnographic analysis of studies included in this review provide a range of barriers to the effective recruitment and retention of ethnic minorities and migrants in community-based interventions in OECD countries. These findings are summarised below.

Barriers related to study characteristics

Studies reported study characteristics contributing to the recruitment and retention of ethnic minorities in RCTs, including intervention types, dose (duration, frequency), and involvement types. For example, studies discussed the different types of interventions aimed at recruiting and retaining participants, such as lifestyle and behavioural interventions [30, 34, 41, 46], interactive group classes on lifestyles [44], nutrition education and counselling by a dietitian [31], curriculum on nutrition and physical activity [9], bicycling [33, 36], exercise [37], behavioural weight management intervention [39], and weight-loss interventions [40]. Types of involvement (physical, face-to-face delivery of interventions [37, 44] telephone and online delivery of intervention [44, 46], also affected participants’ recruitment and retention in RCTs. The duration of the RCTs varied; one for 3 years [2], one for 2 years and 7 months [41], three ran for 2 years [3, 9, 40], one for 18 months [45], and three for 1 year [31, 33, 36]. The rest lasted for less than 6 months [36, 43, 48]. Limited duration of intervention of less than 6 months contributed to decreased participants’ recruitment and retention [46].

Limited access to study sites

Three studies with ethnic minorities, such as, American Indians [10], Latinos [30], and African Americans [31], highlighted the limited access to study site as a reason to not to be part of the recruitment process. Their participation was impacted by long travel distances [10] and transportation problems [2], including limited transport, problems with cars and high travel costs [11, 31, 41].

Time constraints

Time conflict was commonly reported across several ethnic minorities and migrant communities (e.g. Latinos, Brazilian immigrants, African Americans, etc.) [2, 11, 31, 46, 47]. These time conflicts were due to participants’ commitments to family activities, jobs [2, 46], childcare, housing and limited personal time [47], affecting their participation [31]. Participants’ limited availability of time in the afternoon and evening due to children’s after-school activities, work commitments and children’s meals and bedtimes also impacted their participation [46].

Lack of trust

Four studies mentioned a lack of trust as a barrier to effective recruitment and retention of ethnic and migrant communities [2, 10, 30, 35]. According to these studies, participants had a history of distrust with researchers particularly medical care and research, including previous experience of disrespect, poor quality of medical care and researchers not following through commitments. These studies established trust with communities through repeated contacts, listening, shared learning, bi-directional communication, following through on commitments and mutual respect.

Perceived fear and anxiety

Perceived fear and anxiety were a barrier reported in three US studies conducted among Latinos [38], Spanish-speaking adults of Mexican origin [43], and Brazilian immigrants [11]. Many ethnic migrants, such as Mexican Americans [43] and Latinos [38], were afraid of participating in interventions due to fear and anxiety produced by strict anti-illegal immigration laws [43]. Fear of revealing information to researchers also discouraged the participants from taking part in the programmes [11].

Other barriers to recruitment and retention

Other barriers to recruitment and retention included a lack of safety for participants with special needs such as walking difficulty [30], language barriers, low levels of health and intervention literacy [43], and perceived stigma [38]. The current review also found other barriers of participation-associated costs [11], difficulty in maintaining a large pool of potential participants for contact, and prolonged waits in receipt of provider approval [3].

Facilitators of effective recruitment and retention

Facilitators of effective recruitment and retention of ethnic minorities and migrants in RCTs are summarised below. Figures 4 and 5 outline the most effective recruitment and retention strategies, study wise description of these is provided in Supplementary Table 4.

Fig. 4