Results

The literature search resulted in 1180 title and abstract records (Figure 2). Titles and abstracts were largely excluded from full-text review because they were not conducted in the United States, were not conducted in a majority African American population or described an observational study in which authors did not conduct an intervention. Of the sixty-eight full-text articles screened, thirty-five articles met the inclusion criteria.^{(Reference Pierce, Bowden and McCullagh47–Reference Locher, Waselewski and Sonneville80)} Full-text articles were mostly excluded owing to study populations that were not majority African American (n = 18, 54%) or because the studies did not collect participants’ satisfaction with or perceptions of interventions (n = 12, 36%). A small number (n = 3, 10%) were excluded because the studies were observational and lacked an intervention. Following our initial search which resulted in twenty-nine articles, an additional four articles were identified through searching of reference lists,^{(Reference Story, Sherwood and Himes73,Reference Suarez-Balcazar, Martinez and Cox74,Reference Baker, Motton and Seiler78,Reference Williams, Auslander and de Groot81)} and two additional articles were identified in the repeated search in May 2021.^{(Reference Haynes-Maslow, Jones and Morris79,Reference Locher, Waselewski and Sonneville80)} The articles identified in the repeated search included a combined cooking skills and nutrition education programme,^{(Reference Haynes-Maslow, Jones and Morris79)} and a meal delivery programme designed to supplement WIC.^{(Reference Locher, Waselewski and Sonneville80)}

Fig. 2. PRISMA 2009 flow diagram.

Of the thirty-five articles which met our inclusion criteria (Table 1), nearly all (n = 29, 83%) were conducted in urban settings. A minority were conducted in rural settings (n = 3, 8.5%) or settings that were not identified as rural or urban (n = 3, 8.5%). Most studies were conducted with adults (n = 26, 74%), and a smaller number were conducted with youth (n = 6, 17%) or both adults and youth (n = 3, 9%). Most sources included in our review were peer-reviewed journal articles (n = 27, 77%), though some (n = 8, 23%) were dissertations or theses identified through the search of grey literature. No results from the search of webpages met the inclusion criteria.

Table 1. Descriptive characteristics of included studies (n = 30)

A large majority of the included studies examined interventions which consisted of nutrition education alone (n = 22, 63%) (Table 2). For example, several studies assessed participants’ perceptions of nutrition education for school-aged children,^{(Reference Pierce, Bowden and McCullagh47,Reference Shaw53,Reference Baskin, Zunker and Worley55,Reference Blom-Hoffman61)} while others assessed perceptions of educational programs for adults,^{(Reference Ingram49,Reference Davis52,Reference Downes, Buchholz and Bruster54,Reference Jones66,Reference Sheppard, Hicks and Makambi67,Reference Senior Angulo70,Reference Kolavalli71,Reference Conlon, Kahan and Martinez76,Reference Haynes-Maslow, Jones and Morris79,Reference Williams, Auslander and de Groot81)} and some assessed perceptions of education provided by the WIC programme.^{(Reference Joseph63,Reference Cricco-Lizza68,Reference Jackson69)} Five (14.5%) of the studies examined perceptions of combined nutrition education and PSE changes, such as programmes combining a community garden with education.^{(Reference Woodson, Braxton-Calhoun and Benedict58,Reference Grier, Bennette and Covington75)} Another five (14.5%) studies addressed perceptions of environmental interventions alone or in combination with a system intervention (n = 5, 15%), such as the installation of salad bars in majority African American schools.^{(Reference Bean, Spalding and Theriault64)} Finally, several studies evaluated participants’ perceptions of system interventions alone or in combination with a policy intervention (n = 3, 8%), such as an intervention to assess nutritional risk of WIC participants,^{(Reference Dunlop, Dretler and Badal50)} a grocery delivery programme^{(Reference Locher, Waselewski and Sonneville80)} and labelling of calories on restaurant menus.^{(Reference Auchincloss, Chilton and Davis51)}

Table 2. Details of sources addressing African Americans’ perceptions of and/or satisfaction with nutrition interventions

Across all included studies, many collected data regarding attitudes towards or perceptions of an intervention via quantitative surveys alone (n = 10, 28%). In most cases, these quantitative surveys asked participants whether they would attend programmes again or recommend them to a friend. Surveys also asked about specific components of the interventions to assess which portions were most enjoyed and/or accepted by participants. Other studies implemented mixed-methods evaluations, either using quantitative surveys that included open-ended questions or conducting qualitative interviews or focus groups separate from a quantitative survey (n = 7, 20%). Open-ended survey questions tended to focus on suggestions for improvement to the programme and asking which portions of the intervention were preferred by participants. Several studies implemented qualitative methods alone (n = 18, 48%), either through focus group discussions or semi-structured interviews. These qualitative-only studies largely focused on collecting information about participant satisfaction,^{(Reference Pierce, Bowden and McCullagh47,Reference Downes, Buchholz and Bruster54,Reference Cricco-Lizza68,Reference Haynes-Maslow, Jones and Morris79,Reference Locher, Waselewski and Sonneville80)} feedback or suggestions for improvement,^{(Reference Shaw53,Reference Baskin, Zunker and Worley55–Reference Hopkins, Fristad and Goodway57,Reference Grier, Bennette and Covington75)} or individuals’ general perceptions of the intervention.^{(Reference Dunlop, Dretler and Badal50,Reference Auchincloss, Chilton and Davis51,Reference Whitaker, Wilcox and Liu62,Reference Joseph63,Reference Baker, Motton and Seiler78)}

Among studies implementing quantitative methods alone or in combination with other methods (n = 17), twelve evaluated nutrition education interventions alone, two evaluated direct education along with a PSE change and three evaluated PSE changes alone. Among studies implementing qualitative methods alone or in combination (n = 23), fifteen evaluated nutrition education interventions alone, four evaluated direct educations along with a PSE change and four evaluated PSE changes alone. Two studies implemented either phenomenology or ethnography,^{(Reference Jackson69,Reference Kolavalli71)} qualitative methods rarely used to evaluate public health interventions.

Nearly all included publications, whether implementing qualitative or quantitative methods, found that participants had positive experiences with the interventions studied. In general, the included studies found that school children expressed satisfaction with interventions that were perceived as ‘fun’ and had interactive or physical activity components,^{(Reference Pierce, Bowden and McCullagh47,Reference Shaw53)} but found less satisfaction among children with interventions that encouraged them to try new foods,^{(Reference Baskin, Zunker and Worley55,Reference Grier, Bennette and Covington75)} or made changes to the lunchroom.^{(Reference Bean, Spalding and Theriault64)} Studies of educational interventions conducted with adults found that they were perceived to be helpful or useful,^{(Reference Hull, Emerson and Quirk48,Reference Cowart, Biro and Wasserman56,Reference Kannan, Sparks and DeWitt Webster59,Reference Herring, Bersani and Santoro72)} but some participants felt that the interventions were too lengthy or too much of a time commitment.^{(Reference Ingram49,Reference Kannan, Sparks and DeWitt Webster59)} Quantitative surveys found that many participants would recommend the intervention to others,^{(Reference Woodson, Braxton-Calhoun and Benedict58,Reference Kannan, Sparks and DeWitt Webster59,Reference Jones66,Reference Story, Sherwood and Himes73)} found interventions appropriate or acceptable^{(Reference Ingram49,Reference Blom-Hoffman61,Reference Senior Angulo70)} and were satisfied with interventions.^{(Reference Summers and Klassen60,Reference Sheppard, Hicks and Makambi67,Reference Story, Sherwood and Himes73,Reference Suarez-Balcazar, Martinez and Cox74,Reference Williams, Auslander and de Groot81)} Qualitative studies identified participant suggestions for improving interventions, such as shortening lessons,^{(Reference Kannan, Sparks and DeWitt Webster59)} including tips for maintaining success after the programme,^{(Reference Conlon, Kahan and Martinez76)} adding games and food tastings,^{(Reference Grier, Bennette and Covington75)} and increasing interaction among participants.^{(Reference Sheppard, Hicks and Makambi67)} Qualitative studies also identified participants’ perceived benefits of interventions, such as making behaviour changes,^{(Reference Downes, Buchholz and Bruster54)} having greater awareness of health issues^{(Reference Cowart, Biro and Wasserman56)} and using the information learned in lessons in their daily lives.^{(Reference Woodson, Braxton-Calhoun and Benedict58,Reference Kannan, Sparks and DeWitt Webster59)}

Studies examining nutrition education provided through WIC found mixed results regarding satisfaction with the programme. Two qualitative studies found that participants were satisfied with WIC, especially when served by African American WIC staff.^{(Reference Cricco-Lizza68,Reference Jackson69)} In contrast, WIC participants in another publication were dissatisfied with WIC services and saw the programme as intrusive and surveillant.^{(Reference Joseph63)} Only one other study of an educational intervention found that participants were dissatisfied and asserted their own knowledge and expertise related to the intervention topic.^{(Reference Kolavalli71)} All of the studies which found that participants were dissatisfied with interventions implemented qualitative research methods.

Retention rates in intervention studies ranged from 40% to 100% of enrolled participants, and all studies included in the review only solicited feedback from participants who completed the intervention. No studies reported satisfaction or feedback from participants who left the intervention or were lost to follow-up. Few (n = 3) studies reported reasons for participant dropout: one study reported this was due to behavioural or family reasons,^{(Reference Pierce, Bowden and McCullagh47)} and two due to scheduling conflicts.^{(Reference Ingram49,Reference Srivastava, Palmer and Ireland65)}