Outside-of-school time obesity prevention and treatment interventions in African American youth

Transcription

1 obesity reviews doi: /obr Review Outside-of-school time obesity prevention and treatment interventions in African American youth D. J. Barr-Anderson 1, C. Singleton 2,3, C. J. Cotwright 4, M. F. Floyd 5 and O. Affuso 2,3 1 Arnold School of Public Health, Department of Epidemiology and Biostatistics, University of South Carolina, Columbia, SC, USA; 2 Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA; 3 Nutrition Obesity Research Center, University of Alabama at Birmingham, Birmingham, AL, USA; 4 College of Family and Consumer Sciences, Department of Foods and Nutrition, University of Georgia, Athens, GA, USA; 5 College of Natural Resources, Department of Parks, Recreation and Tourism Management, North Carolina State University, Raleigh, NC, USA Received 9 June 2014; accepted 10 June 2014 Address for correspondence: Dr O Affuso, Department of Epidemiology, University of Alabama at Birmingham, Ryals Public Health Building, Room 220E, 1665 University Boulevard, Birmingham, AL , USA. oaffuso@uab.edu Summary Outside-of-school time (OST; i.e. before/after-school hours, summer time), theory-based interventions are potential strategies for addressing increased obesity among African American youth. This review assessed interventions across multiple settings that took place during OST among African American youth aged 5 18 years old. Seven databases were searched for studies published prior to October 2013; 28 prevention and treatment interventions that assessed weight or related behaviours as a primary or secondary outcome were identified. Overall, these studies reported heterogeneous intervention length, theoretical frameworks, methodological quality, outcomes, cultural adaption and community engagement; the latter two attributes have been identified as potentially important intervention strategies when working with African Americans. Although not always significant, generally, outcomes were in the desired direction. When examining programmes by time of intervention (i.e. after-school, summer time, time not specified or multiple time periods), much of the variability remained, but some similarities emerged. After-school studies generally had a positive impact on physical activity, fruit/vegetable consumption and caloric intake, or body composition. The single summer time intervention showed a trend towards reduced body mass index. Overall findings suggest that after-school and summer programmes, alone or perhaps in combination, offer potential benefits for African American youth and could favourably influence diet and physical activity behaviour. Keywords: Health inequity, nutrition, overweight, physical activity. obesity reviews (2014) 15 (Suppl. 4), Introduction Childhood obesity is a significant public health problem given its association with the development of diseases such as type 2 diabetes, hypertension and dyslipidaemia (1). Although results from the through National Health and Nutrition Examination Surveys suggest that obesity rates in children have stabilized, the most recent prevalence estimates indicate that 17.7% of U.S. youth aged 6 11 years and 20.5% of U.S. youth aged years are obese and an additional 16.5 and 14.0%, respectively, are overweight (2). Moreover, racial/ethnic and sex disparities are clearly evident. Non- Hispanic black girls have the highest prevalence of overweight and obesity compared with their peers (2). Although non-hispanic black boys have the lowest rates of overweight and obesity compared with other minority girls and boys, they have experienced the greatest increase in obesity prevalence compared with any other group (8.2% increase in 6 11 year olds and 1.6% increase in year olds from to ) (2,3). These data clearly indicate that successful childhood obesity interventions are still needed, particularly for black/african American youth World Obesity 15 (Suppl. 4), 26 45, October 2014

2 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 27 For more than two decades, school-based interventions have been developed and evaluated to address weightrelated behaviours in youth (4). Schools have been a targeted location because most children and adolescents spend large portions of their days in schools and because schools have health- and physical activity-related mandates (4). However, the time before or after the normal school hours and during the summer (hereafter referred to as outsideof-school time ) is an overlooked and often under-utilized time for weight-related interventions in youth. Recently, several U.S. national and international organizations have promoted targeting weight-related behaviours during outside-of-school time (OST) (5 7). OST interventions can take place in a variety of settings, including schools, community centres, churches and homes. Reviews of school- and community-based physical activity, nutrition and/or obesity interventions have included programmes where results from various settings were combined, regardless of when the intervention was held (4,8). Thus, specific conclusions about OST interventions could not be drawn. Additionally, these reviews did not examine racial/ethnic differences. Some reviews have synthesized the evidence of behavioural interventions targeting weight-related factors that took place during after school (7,9,10). Although 28 unique interventions are presented in the three reviews, one review excluded treatment (i.e. weight reduction) studies (9) and the other three reviews only included studies in which physical activity data were collected (7,10). All of the reviews noted that OST interventions have the potential to positively impact weight-related behaviours among youth. However, none of the reviews examined the impact on the most vulnerable of youth, particularly African Americans. The reviews did not study conclusions specific to this group nor did they provide stratified analysis by race/ethnicity. Furthermore, a search of the literature did not reveal a review of relevant interventions that searched for other outside-of-school times (i.e. other than afterschool), and took place in settings other than schools. Thus, little is known about OST interventions for African American youth. Gaps in the literature related to OST interventions in African American youth remain evident even when referring to more comprehensive reviews that have examined the effectiveness of childhood obesity interventions in minority populations. Seo and Sa (11) reviewed interventions that took place in school, community, home and clinical settings, and included several racial/ethnic minorities (e.g. African American, American Indian, Mexican American, Hispanic and Asian American). Results specific to African American youth were not presented, and some school-based interventions held during regular school hours were also included. Reviews by Baskin et al. (12) and Barr-Anderson et al. (13) included only studies with results specific to African American children and adolescents. Yet, similar to the review by Seo and Sa, interventions conducted during school were not distinguished from OST interventions. However, reviews by Baskin et al. and Barr- Anderson et al. highlighted community engagement and cultural adaptation as potentially important intervention strategies when working with African American youth (14). The core tenets undergirding community engagement include mobilizing groups to identify and use available resources in their own communities, which can influence environmental and behavioural changes to improve the health of community members (15). Very similarly, cultural adaptation is deemed important when engaging racial/ ethnic minorities because framing interventions in the context of the cultural background of the target population can support behaviour change (16). The purpose of the current review was to identify weight-related behavioural interventions for African American girls and boys between the ages of 5 and 18 years that took place during OST. The overarching goal was to survey the childhood obesity literature and to draw conclusions about key intervention components that are relevant when focusing on specific OST periods (i.e. after-school, summer time) and when working with African American youth (i.e. community engagement and cultural adaptation). This review seeks to fill in critical gaps in the literature by addressing the dearth of knowledge about childhood obesity interventions in African American youth, a group at high risk for a myriad of adverse outcomes. It is intended to inform the design of OST interventions to be utilized in the future to address weight-related outcomes in African American youth. Methods We performed a systematic search of studies that evaluated OST interventions targeting weight-related behaviours among African American youth. In September 2013, we searched the following databases for peer-reviewed publications of interventions that aimed to improve health behaviours (i.e. nutrition and/or physical activity or sedentary behaviour) and/or weight status among African American boys and girls aged 5 18 years: AGRICOLA, CINAHL, Cochrane Library, ERIC, NIH RePORTER, PsycINFO and PubMed. Search strategy Search terms were adapted from previous reviews of OST childhood obesity interventions (11,13). These terms included African American or black, intervention or program, child or adolescent, obesity, overweight, weight, BMI, body fat, physical activity, fitness, diet, nutrition, 2014 World Obesity 15 (Suppl. 4), 26 45, October 2014

3 28 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews sedentary, inactivity, after school or after-school, outsideof-school or outside of school, community-based or community based, and home-based or home based. We also examined the reference lists of the retrieved articles to identify other studies that met our inclusion criteria. Inclusion criteria Inclusion criteria were as follows: (i) intervention of 12 weeks or greater in length; (ii) a weight-related behaviour (i.e. diet, physical activity, sedentary behaviour) or weight (obesity status) as a primary or secondary outcome; (iii) study sample 80% African American, or in the case of multi-racial samples, results specific to African American youth available; (iv) intervention conducted OST (i.e. before school, after school, summer time, time not specified, multiple time periods); (v) youth aged 5 18 years as the primary target of the intervention; (vi) intervention included pre- and post-intervention measurements; (vii) published in English language; and (viii) conducted in the United States. Data extraction and analysis Titles and abstracts from the initial search were screened by two research assistants and a co-author (D.J.B.A.) to identify full-length publications that would be reviewed for potential inclusion (Fig. 1). Once a study was identified for inclusion, the co-authors (C.J.C., C.S., D.J.B.A. and O.A.) extracted data pertaining to the inclusion criteria and other key variables using a standardized coding sheet. Detailed quantitative and qualitative information was extracted from each study about health behaviours and obesity status, along with information regarding the study sample and intervention characteristics. Discrepancies in coding were resolved through discussion with the co-authors. All studies meeting the inclusion criteria were summarized based upon study characteristics, intervention strategies and outcome measures/results. The study characteristics were participant characteristics (age range, % female, % African American); study setting (school-, home-, clinic-, or community-based); time of intervention (i.e. before school, after school, summer time, time not specified or multiple time periods); intervention type (prevention or treatment); study design and theoretical framework; and length of intervention including the follow-up period. Both community engagement and cultural adaptation can be implemented at varying degrees in an intervention, which can make it challenging to quantify effectiveness. These intervention strategies were described based upon their relevance when intervening among a racial/ethnic minority group (14). Level of community engagement was defined based upon recommendations from Fawcett et al. (15) and Israel et al. (17) to represent minimal involvement of the community to engagement of the community that aligns with core community-based participatory research Figure 1 Flowchart of systematic review findings. 15 (Suppl. 4), 26 45, October World Obesity

4 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 29 principles: (i) formed community advisory board; (ii) active involvement in design, implementation and study evaluation; (iii) community partners included in authorship; and/or (iv) no community engagement. Degree of cultural adaptation was defined based upon previous reviews (13,18) and described specific adaptations implemented, or stated that adaptations were not specified. The specific adaptations included conducting formative assessment activities (i.e. focus groups, advisory boards) involving African Americans, hiring intervention and/or measurement staff who were African American, and tailoring intervention materials to incorporate culturally relevant content. The methodological quality (MQ) of the included studies was also evaluated. Internal validity was evaluated using six criteria from the Delphi list (19): method of randomization performed, blinding (outcome assessors, interventionists or participants), groups similar at baseline, eligibility criteria specified, point estimates presented and intention-to-treat analyses included. External validity was assessed according to the level of staff expertise, degree to which the programme was adapted to the targeted population, presentation of long-term effects, institutionalization of the intervention beyond the study period and statistical methods for handling missing data (20). The number of internal and external validity criteria met was summed to create a MQ score for a maximum of 11 points. Results related to primary and secondary outcomes were extracted and presented in consideration of MQ and type of study (prevention vs. treatment). When examining outcomes, we were aware that prevention (i.e. weight maintenance) studies tend to have smaller effects on weight than treatment (i.e. weight reduction) studies because prevention studies typically focus on developing and maintaining lifelong healthy behaviours and not weight loss. Additionally, our targeted population included children and adolescents, which can complicate weight, height and body mass index (BMI) change interpretations because they are still growing (21,22). Results Description of studies Table 1 describes the study population, study setting, intervention type, study design, theoretical framework and duration of intervention (and post-intervention follow-up, if applicable) stratified by time of intervention. Thirteen studies occurred after school during an academic school year (23 35), one study took place entirely during the summer (36) and nine studies did not specify the time the intervention was implemented (37 45). No studies were held before school. The remaining five studies conducted sessions partially during the summer and partially during an academic school year. During the school year, these programmes occurred after school (46 48), during an unknown time (49) or any time because the curriculum was available online (50). Twenty of the 28 studies targeted only African American youth (24 27,29 33,35,36,39,43 50), 2 studies had a sample that was greater than 80% African American (23,37) and 6 studies had a sample less than 80% but presented results specific for the African American participants (28,34,38,40 42). Seventeen studies were facilitated in a community (i.e. church, YMCA) (23,25,35, 38 41,44,45,47,49), school (24,26 28,33,34,42) or home setting (37,50). Eight studies were both community- and home-based (29 32,36,43,46,48). Sixteen studies targeted obesity prevention (23 25,27,30 37,44 46,48) and 12 were obesity treatment studies (26,28,29,38 43,47,49,50). The one non-randomized controlled trial (NRCT) (47) and eight uncontrolled trials (UCTs) (23,26 29,35,39,45) were pilot studies. Of the 19 RCTs, 6 were pilot studies (25,30,31,36,38,41) and 13 were full trials (24,32 34,37,40,42 44,46,48 50). The majority of studies (n = 16) cited the use of a theoretical framework, with social cognitive theory (SCT) being the most common; 8 used SCT only (28 31,33,36,47,48) and an additional 3 studies used SCT in combination with another theory (25,34,37). One study used constructs from four different theories (39), and the remaining studies were based upon cognitive behavioural theory (38), stages of change (27), social ecological theory (43) and theory of reasoned action (45). Intervention duration ranged from 12 weeks (23,25,27,30 33,35,36,41,44,45) to 2 years (46,48,50), with the length of follow-up beyond the intervention period ranging from 2 weeks (34) to 2 years (37). Intervention strategies, level of cultural adaptation, degree of community engagement, MQ and study outcomes by time of OST intervention are summarized in Table 2. After-school interventions Approximately 55% of after-school studies focused on changing both physical activity and diet behaviours (25,27,29,31 33,35). Across all after-school studies, the intervention strategies were similar and included face-toface physical activity and/or nutrition behavioural skills sessions. Several programmes offered help with homework (24,30,34,35), assigned take-home assignments (25,32) or incorporated an academic subject such as creative writing (28). The level of cultural adaptation and community engagement ranged from studies describing specific strategies for both (25 27), for neither (23,24,33,34), cultural adaptation only (29 32) and community engagement only (28,35). The seven after-school studies that described cultural adaptations for African American children each 2014 World Obesity 15 (Suppl. 4), 26 45, October 2014

5 30 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews Table 1 Description of outside-of-school time interventions that involved African American children stratified by time of intervention 1 Author and publication year Study population Intervention location Intervention type Study design and theoretical framework Duration of intervention and follow-up After-school 2 Annessi et al (23) Barbeau et al (24) N = years YMCA Prevention Pilot UCT, framework not specified 95% AA 40% female N = 201 Elementary schools Prevention RCT with no 8 12 years 10% female intervention control group, framework not specified 12 weeks, no follow-up 10 months, no follow-up Beech et al (25) N = years Community (location not specified) Prevention Pilot RCT with attention placebo control group, SCT and FST 12 weeks, no follow-up Choudhry et al (26) N = years Community school Treatment Pilot UCT, framework not specified 60% female 14 weeks, no follow-up Engels et al (27) Madsen et al (28) N = 56 Middle school Prevention Pilot UCT, SofC 12 weeks, no follow-up 9 12 years 68% female N = 233 School Treatment Pilot UCT, SCT 32 weeks, no follow-up 8 12 years 18% AA 48% female Resnicow et al (29) N = years Community and participants home Treatment Pilot UCT, SCT 6 months, no follow-up Robinson et al (30) N = years Community centres and participants home Prevention Pilot RCT with attention placebo control group, SCT 12 weeks, no follow-up Story et al (31) N = years School and participants home Prevention Pilot RCT with attention placebo control group, SCT 12 weeks, no follow-up Wadden et al (32) N = years Community clinic and participants home Prevention RCT with head-to-head control group, framework not specified 12 weeks, 6-month follow-up Wilson et al (33) Wilson et al (34) Wofford et al (35) N = 53 Middle school Prevention RCT with attention years 42% female placebo control group, SCT N = 1563 School Prevention RCT with attention years 73% AA 55% female placebo control group, SCT and SDT N = years Community centre Prevention Pilot UCT, framework not specified 39% female 12 weeks, no follow-up 17 weeks, 2 weeks follow-up 12 weeks, no follow-up 15 (Suppl. 4), 26 45, October World Obesity

6 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 31 Table 1 Continued Author and publication year Study population Intervention location Intervention type Study design and theoretical framework Duration of intervention and follow-up Summer time 3 Baranowski et al (36) N = 35 8 years Summer camps and Participants homes Prevention Pilot RCT with attention placebo control group, SCT 12 weeks, no follow-up Time not specified 4 Black et al (37) N = 235 Participant s home Prevention RCT with attention years 97% AA 49% female placebo control group, SCT and MI 11 months, 2-year follow-up Budd et al (38) N = years 43% AA % female not specified Community (location not specified) Treatment Pilot RCT with head-to-head control group, CBT 24 weeks, no follow-up Burnet et al (39) N = years 73% female YMCA and grocery store Treatment Pilot UCT, HBM, SLT, TPB and SET 14 weeks, 1-year follow-up Hasson et al (40); Supplemental article: Davis 2009 (68) N = 100 Community clinic Treatment RCT with no years 48% AA 59% female intervention control group, framework not specified 16 weeks, no follow-up Janicke et al (41) N = years 40% AA 48% female Community (location not specified) Treatment Pilot RCT with attention placebo control group, framework not specified 12 weeks, 6-month follow-up Kang et al (42) N = 80 Academic institute Treatment RCT with no years 68% AA 67% female intervention control group, framework not specified 8 months, no follow-up Naar-King et al (43) N = years 77% female Community and participants home Treatment RCT with attention placebo control group, SET 24 weeks, no follow-up Stolley and Fitzgibbon 1997 (44) N = 65 Tutoring centre Prevention RCT with attention 7 12 years placebo control group, framework not specified 12 weeks, 1-year follow-up Thompson et al (45) N = 39 Church Prevention Pilot UCT, TRA 12 weeks, no follow-up years After-School and Summer Time 5 Klesges et al (46) N = years Community centres and participants home Prevention RCT with attention placebo control group, framework not specified 2 years, no follow-up Raman et al (47); Supplemental article: Ritchie 2010 (69) N = 165 YMCA Treatment Pilot NCRT with 9 11 years 56% female attention placebo control group, SCT 12 months, no follow-up 2014 World Obesity 15 (Suppl. 4), 26 45, October 2014

7 32 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews Table 1 Continued Author and publication year Study population Intervention location Intervention type Study design and theoretical framework Duration of intervention and follow-up Robinson et al (48) N = years Community centres and participants home Prevention RCT with attention placebo control group, SCT 2 years, no-follow up Time not specified and summer time 6 Resnicow et al (49) Any time and summer time 7 Williamson et al (50) N = 123 Churches Treatment RCT with years head-to-head control group, framework not specified N = 57 Participants home Treatment RCT with attention years placebo control group, framework not specified 6 months, 6-month follow-up 2 years, no follow-up 1 Time of intervention refers to when the primary target of the study (child) participated in the outside-of-school time intervention (i.e. after school during the school year, during summer time, time not specified or multiple time periods). 2 Interventions occurred after school during academic school year. 3 Interventions were implemented during summer time only and not during academic school year. 4 When the intervention took place was not specified. However, authors indicated that the programme did not occur during normal school hours. 5 Interventions spanned at least one full calendar year (January December) which includes both academic school year and summer time. During the academic school year, the intervention was implemented after school. 6 Intervention was implemented partially during the academic school year and partially during the summer. During the academic school year, the time was not specified. 7 Intervention was Internet-based; therefore, content was available 24 h/d 7 d/week. Additionally, the intervention spanned a full calendar year which included summer. AA, African American; CBT, cognitive behavioural theory; FST, family systems theory; HBM, health belief model; MI, motivational interviewing; NRCT, non-randomized controlled trial; RCT, randomized controlled trial; SDT, self-determination theory; SCT, social cognitive theory; SET, social ecological theory; SLT, social learning theory; SoC, stages of change; TPB, theory of planned behaviour; TRA, theory of reasoned action; UCT, uncontrolled trial. reported culturally tailoring the content of intervention materials and messages (25 27,29 32). Two studies also utilized African American instructors to deliver the intervention (30,31) and one after-school study added formative assessment (i.e. focus groups with African American children and/or parents) to guide the study strategies (49). Although most after-school studies did not engage the community, the five studies that did engage the community reported actively involving the community in the design, implementation and evaluation of the study (25 28,35). MQ for the after-school studies ranged from 1 to 9 out of a possible score of 11 (Table 2), with a mean MQ score of 4.4 (±2.6). Additional details on specific ratings for each criterion can be found in Appendix 1. Furthermore, study outcomes were evaluated in the context of MQ and type of study, both of which can influence the interpretation of results (Table 2). Ten of the 13 after-school studies were prevention-focused. Of the six studies that utilized a RCT study design, significant beneficial changes in physical activity (24), screen time (30), fruit and vegetable intake (33), and BMI and percent body fat (24) were reported. Although prevention RCTs produced outcomes generally in the desired direction, only the aforementioned studies with MQ scores of 9 (30), 6, 5 (24), 8 and 3 (33) reported statistical significance. Among the four uncontrolled prevention studies, two reported positive outcomes for body composition measures (23,27), with one reaching statistical significance (23). The remaining two uncontrolled prevention studies reported stability in BMI from baseline to the post-intervention assessment (35,45). Significant positive effects were also noted for the behavioural outcomes in two of the uncontrolled prevention interventions (27,35). Specifically, Engels et al. reported an increase in fruit consumption, while Wofford et al. found reductions in sugar-sweetened beverages and screen-time and an increase in physical activity. However, the MQ scores for these studies were among the lowest: 3 (27,45), 2 (23) and 1 (35). The three after-school studies that targeted treatment were uncontrolled and reported significant reductions in BMI z-score (26,28) and a non-significant reduction in percent body fat (29). Results were mixed for the 15 (Suppl. 4), 26 45, October World Obesity

8 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 33 Table 2 Intervention strategies, cultural adaptation, community engagement and outcome results stratified by time of intervention Author and publication year Specific intervention strategies Cultural adaptation Community engagement Methodological quality, study type and design Results 1 After school Annessi et al (23) A 3 d/week PA curriculum that included cardiovascular activities, resistance exercises and behavioural skills training Behavioural skills training included interactive lessons on goal setting, progress monitoring, facilitative self-talk and identifying appealing physical activities Not specified Not specified 2, prevention UCT *%BF: + Barbeau et al (24) After-school programme that included homework, healthy snack time and physical activity (25-min PA skill instruction, 35-min aerobic PA such as basketball, tag, softball, relay races and 20-min strengthening/stretching) Incentives for attendance Not specified Not specified 5, prevention RCT *BMI: + WC: + *%BF:+ *VAT:+ *MVPA:+ VPA: + Beech et al (25) Physical activity (hip hop aerobics) sessions Healthy eating session with taste-tests of healthy foods and food preparation/games Culturally relevant take-home materials Culturally sensitive programming and tailored take-home materials Active involvement in design, implementation, and evaluation of study 8, prevention RCT BMI: 0 WC: + SSB: + PA: + Choudhry et al (26) After-school programme that consisted of 14, 1-h weekly sessions on nutrition and PA Hand-on activities including recipe development, role playing, dance and art projects Open group discussions Environmental change in school that included displays, showcases and art contests related to health education Formative assessment with AA parents. Culturally tailored curriculum that incorporated Afrocentricism and other AA values Active involvement in design, implementation and evaluation of study 3, treatment UCT *BMIz: + Engels et al (27) min sessions 4 d/week Dance, sport games and other fitness activities Pedometers provided Targeted educational handouts on nutrition and fitness Recording of fruit and vegetable intake and step counts Poster board displays in school Culturally tailored PA activities (use of African dance) Active involvement in design, implementation and evaluation of study 3, prevention UCT BMI: %BF: + *Fruit: + Vegetable:+ PA: Madsen et al (28) Soccer (2 practice days/week with up to 2 h of MVPA; 1 game day with warm up period followed by 1 h game) Community service projects (2 other days/week such as community food drives, neighbourhood beautification) or creative writing Not specified Active involvement in design, implementation, and evaluation of study. Community partners included in authorship 1, treatment UCT *BMIz: OW/OB BMIz: 0 Fitness: World Obesity 15 (Suppl. 4), 26 45, October 2014

9 34 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews Table 2 Continued Author and publication year Specific intervention strategies Cultural adaptation Community engagement Methodological quality, study type and design Results 1 Resnicow et al (29) Sessions included min of PA, meal preparation lessons and an interactive educational activity Educational activities included reading food labels and learning about satiety Field trips to grocery stores Step aerobics, commercial aerobics videos, toning, walking and outdoor games led by staff Culturally tailored PA programming Not specified 4, treatment UCT BMI: %BF: + Fitness: kcal: + Robinson et al (30) After-school dance classes with healthy snack, homework period and discussion of increased physical activity (dance) and reduced TV screen time (TV watching, videotape use and video game use) Family intervention which included role modeling for girls by African American interventionist and behaviour change discussions about reducing screen time Newsletters AA only interventionists and data collectors. Attempted to account for unique elements associated with AA culture Not specified 9, prevention RCT BMI: + WC: + PA: + *TV: + Story et al (31) Physical activity and healthy eating, behavioural skills programme based upon youth development and resiliency approach to build on family and personal strengths Family night events with interactive games and goal setting that they would continue throughout the programme Phone calls by staff to check in on goals and provide support AA only instructors. Culturally tailored activities and programming Not specified 5, prevention RCT BMI: WC: MVPA: + kcal: + Wadden et al (32) Incentive structure based upon weight loss and attendance Take-home assignments Various levels of parental involvement based on treatment condition (parents received homework assignments, participated in programme with girl, or talking with daughter or attended separate similar session) AA-adapted curriculum content Not specified 6, prevention RCT BMI: + Wilson et al. Nutrition and PA education sessions 2002 (33) 2 Video created that focused on positive coping strategies to increase healthy lifestyle behaviours Not specified Not specified 3, prevention RCT SCT group only *Fruit & vegetable: + kcal: PA:+ SCT + MI group only *Fruit & vegetable: + kcal: + PA: 15 (Suppl. 4), 26 45, October World Obesity

10 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 35 Table 2 Continued Author and publication year Specific intervention strategies Cultural adaptation Community engagement Methodological quality, study type and design Results 1 Wilson et al (34) Snack and homework help Non-competitive play (five-pass basketball) for 1 h of MVPA; 3 d/week Sessions 3 d/week on behavioural skills and motivation for increasing PA at home Practice at home 1 d/week Not specified Not specified 7, Prevention RCT MVPA: + Wofford et al (35) After school programme with nutritious snacks and homework help Health lessons (nutrition, increasing physical activity, and relationship between nutrition intake and physical activity) Active activities dancing Team building activities Not specified Active involvement in design, implementation, and evaluation of study 1, Prevention UCT *Screen-time: + *PA: + *SSB:+ BMI%tile: 0 Summer time Baranowski et al (36) Camp programme to increase behavioural and psychosocial factors related to healthy foods (i.e. fruit and vegetable intake, water consumption) and physical activity Self-monitoring using pedometers Goal-setting website Formative focus groups with AA sample Not specified 6, prevention RCT WC: OW/OB BMI: + Time not specified Black et al. Role modelling and support from AA college mentors 2010 (37) 3 Goal setting related to PA and diet Healthy snack preparation (taste tests, recipe sharing) Culturally tailored diet and PA activities. Formative assessment assisted by advisory board of AA adolescents Formed a community advisory board of AA adolescents 8, prevention RCT OW/OB group only *PA:+ *BMIz:+ *%BF:+ *FM:+ *FFM: + Budd et al (38) Comprehensive family-based intervention that included 13 group sessions and 6 bimonthly sessions Parents and children received health education and were encouraged to consume 1,200 1,500 kcal/d and exercise 120 min/week Not specified Not specified 6, treatment RCT WT: + BMI: + BMIz: + *WC: + Burnet et al (39) Physical activity and nutrition discussion topics Behavioural goal setting Skill building and group problem solving Engaging family activities (shopping, cooking, exercise) Self-monitoring practice Group outings Formative focus groups with AA families. All female AA lay community leaders Participated in formative focus groups 6, treatment UCT *BMIz: + *WHR: %BF: + VPA: Walk: + TV: 2014 World Obesity 15 (Suppl. 4), 26 45, October 2014

11 36 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews Table 2 Continued Author and publication year Hasson et al (40) 4 ; Supplemental article: Davis 2009 (68) Janicke et al (41) Kang et al (42) Naar-King et al (43) 5 Specific intervention strategies Cultural adaptation Community engagement Methodological quality, study type and design Results 1 Modified carbohydrate nutrition education programme once a week Strength training twice a week Culturally tailored nutrition programme Not specified 4, treatment RCT NUTR group LB strength: + *kcal: CHO: +, *Added sugar: + *FM: NUTR+ST group *LB&UB strength:+ *kcal: +, *Added sugar: + Knowledge and skill-based education sessions Food and pedometer logs (pedometers provided) Group support meetings (separate parent/child meetings for learning component and together for goal-setting Taste-testing and snack prep for children Exercise or games for children to be active Not specified Not specified 4, treatment RCT BMIz: 1-h lifestyle education class every 2 weeks for 8 months (class covered: principles of learning and behaviour modification, information about nutrition and physical activity, discussions of various aspects of the food consumption process, review of psychosocial factors related to obesity, and problem-solving/coping skills) Lifestyle education + moderate intensity physical training 5 d/week for 8 months Monetary incentives for attending sessions Not specified Not specified 5, treatment RCT %BF: + VAT: + *Fitness: + Multisystemic therapy intervention that involved children and parents Used targeted techniques to assess barriers at the individual level (i.e. beliefs about foods and managing cravings), the family level (i.e. parental monitoring and teaching nutrition using foods in home), and the extra-familial level (i.e. exercising with peers and engaging school personnel) Not specified Not specified 8, treatment RCT Intervention group only *%OW: + *%BF: + BMI: + 15 (Suppl. 4), 26 45, October World Obesity

12 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 37 Table 2 Continued Author and publication year Specific intervention strategies Cultural adaptation Community engagement Methodological quality, study type and design Results 1 Stolley and Fitzgibbon 1997 (44) Nutrition education sessions Cooking demonstration Music and dance incorporated into nutrition and physical activities Culturally tailored content and programming Not specified 6, prevention RCT *% dietary fat: + Thompson et al (45) Physical activity log Aerobic dance class Physical activity education (knowledge about PA, goal setting, benefits and barriers, body image, role models, social support, hair maintenance, health statistics, solicit feedback from girls about changing environments) AA only interventionists. Incorporated AA cultural values in intervention Not specified 4, prevention UCT BMI: *Fitness: After school and summer time Klesges Monthly field trips et al Nutrition and physical activity sessions (goal setting, provided (46) 6 feedback, encouragement to participants, skill building, self-monitoring, problem solving, and social support) Parents/guardians were encouraged to make changes in the home food environment Culturally sensitive programming and tailored take-home materials Not specified 9, prevention RCT YR1 results BMI: *WC: %BF: + YR2 results: BMI: + WC: %BF: + Raman et al (47); Supplemental article: Ritchie 2010 (69) Summer day camp with community-based exercise, nutrition and behavioural modification Monthly nutrition educational sessions Personal best approach to physical activity programming to create an environment where overweight children develop positive self-esteem and respect Follow-up: Weekly intervention sessions including physical activity and modeling, hands-on nutrition education and skill-building, and self-esteem modelling Outside-of-programme physical activity Not specified Not specified 3, treatment NRCT BMIz: %BF: WT: + WC: + Robinson et al (48) Afterschool programme with dance, healthy snack, and homework Public performances START (Sisters Taking Action to Reduce Television) home-based screen time reduction programme (self-monitoring, a 2-week TV-turnoff, budgeting viewing hours, intelligent viewing) Newsletters AA culture infused in intervention activities Community participated in formative focus groups 8, prevention RCT BMIz: WC: + TSF: + PA: + Screen time: + kcal: World Obesity 15 (Suppl. 4), 26 45, October 2014

13 38 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews Table 2 Continued Author and publication year Specific intervention strategies Cultural adaptation Community engagement Methodological quality, study type and design Results 1,2 Time not specified and summer time Resnicow 30 min of physical activity et al Taste-test and preparation of healthy foods (49) Dependent on treatment group, motivational interviewing counselling calls Retreat at national park Two-way pagers for targeted messages Formative assessment with AA families Community participated in formative focus groups 7, treatment RCT BMI: + WC: + %BF: + Fitness: Any time and summer time Williamson Internet-based, interactive nutrition education and counselling et al via intervention website/ (50) Face-to-face counselling session Behavioural self-monitoring online log Culturally tailored activities (i.e. common AA recipes, links to AA health web sites). Counsellors educated on culturally specific health information and diet/pa-related issues Not specified 6, treatment RCT %BF: + BMI: *Indicates a statistically significant difference at a level of P < Interpretation of outcome results: For randomized controlled trials and non-randomized controlled trials, outcomes reported are for between group differences unless denoted; a plus sign (+) indicates a treatment minus control difference in the desired direction and a minus sign ( ) indicates a difference opposite to the desired direction. For uncontrolled trials, outcomes reported are for within-group differences unless denoted; a plus sign (+) indicates a post-intervention minus baseline difference in the desired direction, a minus sign ( ) indicates a difference opposite to the desired direction, and a zero (0) indicates no change. 2 Results for Wilson et al. (33) study are presented for within-group differences for SCT (social cognitive theory-based intervention) and SCT + MI (social cognitive theory plus motivational interviewing-based intervention) groups. 3 Results for Black et al. (37) study are presented for the overweight/obese (OW/OB) group only. 4 Results for Hasson et al. (40) study are presented for within-group differences for NUTR (nutrition education only intervention) and NUTR + ST (nutrition plus strength training intervention) groups. 5 Results for Naar-King et al. (43) are presented for the intervention group only. 6 Results for Klesges et al. (46) study are presented for 1-year (YR1) and 2-year (YR2) follow-up. %tile, percentile; AA, African American; BF, body fat; BMI, body mass index; BMIz, body mass index z-score; CHO, cholesterol; FFM, fat-free mass; FM, fat mass; kcal, kilocalories; LB, lower body; MQ, methodological quality; MVPA, moderate/vigorous physical activity; NS, not specified; OB, obese; OW, overweight; PA, physical activity; ST&D, study type and design; SSB, sugar-sweetened beverage; TSF, triceps skinfold; UB, upper body; Walk, >2 h walking a day; WC, waist circumference; WHR, waist-to-hip ratio; WT, weight; VAT, visceral adipose tissue; VPA, vigorous physical activity. 15 (Suppl. 4), 26 45, October World Obesity

14 obesity reviews Obesity interventions in African American youth D. J. Barr-Anderson et al. 39 behavioural variables. Madsen et al. reported positive changes in fitness (28), in contrast to Resnicow et al. s negative findings for fitness but positive results for caloric intake (29). Both of these studies had MQ scores less than 3. Although there was some variance in results across study design, in general, after-school studies positively impacted physical activity, fruit/vegetable consumption and caloric intake, and body composition. Intervention implemented during summer time only One OST intervention was conducted exclusively during the summer time (36). This intervention targeted nutrition and physical activity behaviours by combining summer day camp activities with a website for goal setting. The primary strategies included skill-building to facilitate increased fruit, vegetable and water intake as a means of replacing excess calories from high-fat foods and soft drinks, as well as to increase physical activity through the use of pedometers. Although community engagement was not reported, focus groups with African American girls and parents led to the cultural adaptation of using a summer day camp vs. live-in camp model for the study. This pilot prevention RCT (MQ score = 6 of 11) tended towards a reduction in BMI among overweight/obese African American girls. Unspecified time of intervention Among the nine studies where time of intervention was not specified, all (37 45) but 1 (45) targeted both nutrition and physical activity behaviours. Thompson et al. focused solely on increasing physical activity (45). Eight (38 45) of the nine studies were conducted in a group setting, and family/parent participation was reported in four (38,39,41,43) of the nine studies. A wide range of intervention strategies were used among these studies, including food demonstration and preparation (37,39,44), caloric or carbohydrate restriction (38,40), dance (44,45), goal setting (37,39,41) and parental monitoring (43). Five of the nine studies included some level of cultural adaptation, including focus groups (37,39), cultural tailoring of activities (40,44,45) and African American intervention staff (39,45). However, only two of the nine studies specified community engagement, which included a community advisory board (37) and interviews with community leaders (40). The mean MQ score of studies without time specified was 5.7 ± 1.6 of 11 possible points. Three studies in this category were prevention trials, including two RCTs (37,44) and one UCT (45). Black et al. reported statistically significant positive effects on both the physical activity and the body composition outcomes among overweight and obese participants (37); this study had an MQ score of 8. Stolley and Fitzgibbon, whose study had a MQ score of 6, found statistically significant reductions in dietary fat postintervention (44). In contrast, fitness was negatively impacted in the UCT (MQ score of 4), although BMI was stable from baseline to post-intervention (45). The remaining six studies were treatment trials, including five RCTs (38,40 43) and one UCT (39). Two of the five treatment RCTs with relatively high MQ scores reported significant reductions in weight-related outcomes (38,43), whereas two other RCTs with relatively low MQ scores produced results in the expected direction for weight-related and behavioural outcomes (e.g. increase in fitness) (42) or no change in BMI z-score (41). However, Hasson et al. reported significant results for fat mass loss among the nutrition education group, whereas no change in fat mass was reported for the nutrition education plus strength training group (40). The sole treatment UCT, which had a MQ score of 6, reported a significant reduction in waist-to-hip ratio, but non-significant reductions in TV time and vigorous physical activity (39). Interventions implemented during multiple time periods: unspecified times and summer time Resnicow et al. conducted a church-based intervention during the school year and summer time that focused on improving nutrition and physical activity behaviours (49). However, the time of the intervention was not specified. Intervention strategies included physical activity sessions, healthy food preparation, motivational interviewing and targeted messaging via two-way pagers. Cultural adaptation and community engagement were evidenced by formative assessments with African American families and participation in the study design by church representatives, respectively. The MQ score for this treatment RCT was 7; weight-related results were in the expected direction, and the fitness outcome was in the negative direction. Interventions implemented during multiple time periods: anytime and summer time Williamson et al. conducted a 2-year Internet-based intervention that allowed for continuous access to intervention materials during the study period (50). The treatment group website strategies targeted nutrition and physical activity behaviour change; the control group website provided minimal health education information. However, both groups received limited face-to-face counselling sessions. The intervention employed cultural adaptation in the form of culturally tailored websites and training for counsellors. However, no community engagement was specified. This treatment RCT with a MQ score of 6 resulted in 2014 World Obesity 15 (Suppl. 4), 26 45, October 2014

15 40 Obesity interventions in African American youth D. J. Barr-Anderson et al. obesity reviews percent body fat changes in the expected direction, but BMI in the negative direction, although neither reached statistical significance. Interventions implemented during multiple time periods: after school and summer time Two of the three after-school interventions that included programme activities during the summer time targeted the improvement of both nutrition and physical activity behaviours (46,47). One remaining study in this category focused on increasing physical activity and reducing screen time (48). A variety of strategies were used by these year-round interventions such as sessions on self-monitoring (46 48), monthly field trips (46), recommendations for changes to the home environment (46), hands-on nutrition education (47), self-esteem modelling (47), dance (48) and TV time budgeting (48). Two of the year-round interventions included cultural adaptation techniques (46,48). Klesges et al. developed culturally tailored take-home messages (46), whereas Robinson et al. incorporated both culturally tailored programming and African American cultural values into the intervention (48). The latter study included community focus groups as a form of community engagement (48). The two year-round prevention RCTs that were conducted both after school and during summer time had relatively high MQ scores but no statistically significant outcomes (46,48). However, Klesges et al. reported weightrelated outcomes in the expected direction postintervention for younger girls (46), whereas Robinson et al. reported increases in BMI z-score and decreases in waist circumference (48). Furthermore, the single year-round treatment non-rct included in this review yielded no significant findings for any of the weight-related outcomes (48) and had a low MQ score of 3. Discussion The purpose of this review was to identify and critically assess obesity prevention and treatment interventions for African American girls and boys between the ages of 5 18 years that took place during OST. With obesity affecting so many African American children and adolescents, it is well understood that effective interventions are needed. The current review synthesized the evidence of interventions that occurred during OST, a prime, but possibly underutilized temporal setting. We examined the 28 interventions included according to the specific OST frame (i.e. after school, summer time, time not specified, multiple time periods) in which the interventions were implemented: 13 occurred after school, 1 during the summer time, 9 during unspecified times and the remaining 5 during multiple time periods. Across time of intervention, diversity in implemented intervention strategies, the degree of cultural adaptation and the level of community engagement were evident among the studies. The inconsistency in MQ scores, imbalance of full trials vs. pilot studies and variability of study designs among the interventions made it challenging to draw overarching conclusions about effective strategies to address excess weight, physical activity or healthy eating in these minority youth. Similar inconsistencies have also been problematic for other review articles (13,51,52). Generally, this review supports a positive impact of OST interventions that focus on either prevention or treatment of weightrelated factors with African American youth. Stratifying the interventions based upon when they were offered allowed for further conclusions to be drawn based upon timing. The after-school interventions included face-to-face, physical activity and/or nutrition behavioural skills sessions with many incorporating an academic activity (i.e. homework help, take-home assignments or creative writing). There was no consistent pattern of cultural adaptation or community engagement for these programmes. However, regardless of the study design, after-school studies tended to positively impact physical activity, fruit/vegetable consumption and caloric intake and body composition. These interventions also reported more educational or academic activities than interventions that took place during a different time frame. This may have been due to the fact that many of the after-school interventions were associated with a school or an already established learning-based after school programme. Additionally, these programmes took place during a time when students usually complete homework or may engage in some academic activity. Physical activity and good nutrition have been positively associated with academic achievement (53,54); after school interventions that promote both healthy behaviours and enhanced learning have the potential for positive impact in multiple ways. This is especially important for African American youth because of the achievement gap (55) and weightrelated health disparities (2) between this population and white youth. It appears that a culturally adapted summer time intervention in a day camp setting may be beneficial in changing nutrition and physical activity behaviours among African American girls. Although attendance rates were high in both the intervention and the control groups (>80% daily attendance) (36), the length of the active intervention may need to be greater than 4 weeks prior to transitioning to a minimal Web-based intervention strategy to increase the likelihood of significant changes in weight-related outcome. It was not clear whether or not the majority of summer camp activities took place outside. However, the intervention schedule suggests that outdoor activities (i.e. soccer, pool day) may have been included. Overall, the findings of 15 (Suppl. 4), 26 45, October World Obesity