Cultural Adaptation for Ethnic Diversity : A Review of Obesity Interventions for Preschool Children

Obesity disproportionately affects U.S. ethnic minority preschool children, placing them at risk for obesity related co-morbidities and premature death. Effective culturally appropriate interventions are needed to improve health behaviors and reduce obesity in young high-risk minority children, while their behaviors are still developing. All known obesity intervention studies (e.g., diet and physical activity) since 2000 targeting U.S. ethnic minority preschool children were reviewed. Five electronic databases and eight published literature reviews were used to identify the studies. Intervention studies without identified ethnic minority participants were excluded. Ten obesity interventions studies met the review criteria. Published cultural adaptation guidelines were used to develop a mechanism to analyze, score, and rank the intervention adaptations. Cultural adaptations varied widely in rigor, depth, and breadth. Results indicated a relative absence of appropriately adapted obesity interventions for ethnic minority groups, suggesting a need for more rigorous cultural adaptation guidelines when designing obesity interventions for diverse ethnicities. Culturally appropriate adaptations appeared to enhance intervention relevance, effectiveness, and feasibility. The purpose of this literature review was to evaluate 1) the type and extent of cultural adaptations strategies applied to the interventions, and 2) how these adaptations related to the study outcomes. © 2011 Californian Journal of Health Promotion. All rights reserved.


Introduction
The obesity epidemic is a major public health concern in the United States.One-third of all children are overweight or at risk for becoming overweight.Of these children, one-fourth are toddlers and preschoolers (2-to 5-years old) (Ogden, Carroll, & Flegal, 2008).All face possible deteriorating health from cardiovascular disease and diabetes, leading to premature death (Franks et al., 2010;Goran, Lane, Toledo-Corral, & Weigensberg, 2008).Among preschool children, a disparity in the prevalence of obesity is especially evident in low-income, ethnic groups (Anderson & Whitaker, 2009;Division of Nutrition, Physical Activity, and Obesity, 2009).Alaskan Indian/Native American preschool children are at highest risk for obesity, followed by Hispanics and non-Hispanic blacks, compared to non-Hispanic white and Asian children.
A promising strategy to help reduce childhood obesity is to instill healthy behaviors in high-risk preschool children while they are young and their lifestyle behaviors are still developing (Kimbro, Brooks-Gunn, & McLanahan, 2007;Nader et al., 2006).Recent systematic reviews found limited obesity intervention studies focused on ethnic minority preschool children (Branner, Koyama, & Jensen, 2008;Brown, Kelly, & Summerbell, 2007;Wilson, 2009).The U.S. Surgeon General (U.S.Department of Health and Human Services [DHHS], 2010) and the Institute of Medicine (IOM) (2006) call for more research to identify effective interventions for groups at high risk for obesity.Additional recommendations are for research designs to be more culturally appropriate for diverse ethnic minorities.

Objective
The purpose of this literature review was to evaluate cultural adaptations made to obesity intervention studies targeting high-risk U.S. ethnic minority (e.g., Native Americans, Hispanics, and/or African American) preschool children (2-to 5-year-olds).This review addressed the extent to which cultural adaptation strategies were present in each intervention and how these strategies related to study outcomes.
The importance of culturally relevant interventions and suggestions for practical adaptation strategies to improve health promotion interventions are also addressed.Adaptations may help improve the quality of health promotion programs for ethnic minority preschool children and their families and address the growing demand for guidance in adapting interventions to effectively influence healthy behaviors (Cluss, Ewing, Long, Krieger, & Lovelace, 2010;Sanders Thompson et al., 2008).
The literature search found no published literature reviews evaluating cultural adaptation of obesity intervention studies for high-risk U.S. ethnic minority preschool children.This deficiency highlights the need for reviews evaluating cultural adaptations of obesity intervention studies for this population group.

Need for Cultural Adaptation of Interventions and Measures
When designing interventions, it is important to take into account the unique cultural values, beliefs, socio-economic status (SES), and environment of ethnic minority populations (Elder, Ayala, Parra-Medina, & Talavera, 2009;Hurst & Nader, 2006).Improving health behaviors may be less effective if investigators disregard the need for cultural adaptation or inadequately adapt interventions (Marin, 2006).Some ethnic groups may perceive culturally inappropriate interventions as confusing, irrelevant, and/or offensive (Castro, Barrera, & Martinez, 2004;Marin, 2006), resulting in less participant engagement, compliance, and retention.
Given the wide variability within and among cultural groups, investigators may not have the cultural competence to effectively adapt an intervention for a particular ethnic group.For example, African American and Hispanic designations consist of multiple distinct subgroups defined not only by race, but also by regional, national, and continental origin (Kreuter, Lukwago, Bucholtz, Clark & Sanders-Thomason, 2003;Office Of Minority Health, 2009), such as Puerto Ricans and Columbians.Clinicians and investigators have expressed limited awareness of published guidelines for culturally adapting interventions and translating materials (Cluss et al., 2010;Sanders Thompson et al., 2008), supporting the need for standardized guidelines.
Development of culturally equivalent versions of original measures (e.g., surveys, questionnaires, and interview guides) is needed to accurately determine the effectiveness of interventions with different cultural groups (Castro et al., 2004).Simply translating a measure verbatim into the ethnic group's dominant language is insufficient for adapting key constructs, concepts, and content, thus invalidating the measure (Ramirez, Ford, Stewart, & Teresi, 2005;Martinez, Ainsworth, & Elder, 2008).Diverse ethnic populations differ in cultural perspectives, and many original constructs and concept measures may not be commonly shared or understood unless they are appropriately translated (Castro et al., 2004).
Translated instruments should be: (a) equivalent to the original instrument; (b) culturally, conceptually, and contextually relevant for the intended audience; and (c) reliable and valid.Invalid measures (do not reflect original concepts) and unreliable instruments (inconsistent across settings) (Creswell 2009) may bias results, leading to irrelevant, ineffective, and financially wasteful policies and health services for ethnic minority groups (Davidson & Knafl, 2006;Ramirez et al., 2005).
Surface versus deep structure.Resnicow, Baranowski, Ahluwalia, and Braithwaite (1999) conceptualized cultural sensitivity in two dimensions -surface structures and deep structures.Surface and deep structure adaptation are strategies used to design culturally and linguistically relevant measurement instruments, interventions, and program materials.Surface structure adaptations employ visual and auditory cues for culturally appropriate messages reflecting settings and lived experiences of the ethnic group, including: music, pictures, foods, clothing, locations, and people.
Deep structure adaptation is more abstract and usually more difficult to achieve than surface structure adaptations because it is complicated and time consuming.It involves cultural sensitivity and a comprehensive understanding of the ethnic group's core cultural values, norms, and stressors (e.g., economic, social, and environmental) affecting health behaviors (Resnicow et al, 1999).For example, it may be prudent to focus on health rather than obesity to avoid stigmatizing overweight participants or offending members of cultures that favor larger body builds (Tailor & Ogden, 2009).Deep structure adaptation may require input from the community and can dramatically improve the effectiveness of health promotion interventions.
Targeting and tailoring interventions.Additional cultural adaptation concepts include targeting and tailoring to address how broad or focused an intervention should be to affect change in the intended population.Targeting is a broadly tuned intervention approach designed to reach most group members and presumes sufficient population homogeneity.In contrast, tailoring creates a more fine-tuned intervention to reach one specific person or subgroup (Elder, Ayala, Slymen, Arredondo, & Campbell, 2009;Kreuter et al., 2003).
Categories of cultural adaptation strategies.To effectively influence healthy behavior changes, culturally adapted interventions may require modifications tailored to a target group's worldview (Elder, Ayala, Slymen, et al., 2009).Depending on an ethnic group's characteristics, different cultural adaptation strategies may be required to modify interventions and program materials.Krueter and associates (2003) organized commonly used intervention adaptation strategies into five categories: (a) peripheral, (b) evidential, (c) constituentinvolving, (d) socio-cultural, and (e) linguistic.
Peripheral strategies target a culture's surface structure by incorporating audio and/or visual elements, (e.g., music, colors, pictures, clothing, ethnic foods, or people) easily recognized by members of the ethnic group.Evidential strategies present scientific evidence regarding health issues relevant to the intended audience (e.g., causes and prevalence of certain illnesses or diseases common in the target population).For example, noting that Hispanic children are at high risk for Type 2 diabetes is an evidential strategy.
Constituent-involving strategies solicit participation, knowledge, and input from members of the target community (e.g., lay health workers, leaders, and focus groups) regarding the culture's deep structure (e.g., cultural norms, beliefs, social structure, and SES).Community participation is essential to help develop a culturally appropriate intervention program (Horn, McCracken, Dino, & Brayboy, 2008).Community collaboration allows investigators to develop cultural sensitivity, employ local cultural resources, understand community health issues, and solicit stakeholders' input in the intervention design (Cardona et al., 2009).This is useful for incorporating both surface and deep structure within an intervention.Socio-cultural strategies incorporate extensive social and cultural values into the intervention design to provide context and meaning or "deep structure" for the intended group.Examples of socio-cultural strategies might include seeking family approval when recruiting participants from Hispanic cultures or providing child-care and social services for participants with limited finances (Resnicow et al., 1999).Making the intervention meaningful and relevant for participants can promote participant engagement and retention, which may improve program effectiveness.
Linguistic strategies typically follow established guidelines to develop culturally equivalent translations of measurement instruments, consent forms, and so on.The intent of linguistic strategies is to retain the concepts and constructs of the original materials and instruments while using language easily understood by the target population (Ramirez et al., 2005).Most published translation methods follow similar guidelines using forward and backward translation, independent bilingual translators, and multiple reviewers.
For example, Eremenco, Cella, and Arnold (2005) introduced a universal translation method for program materials and instruments.This method attempts to limit measurement bias from self-reported health measures by producing an equivalent translation of the original instrument or materials.It has been successful in retaining original concepts and constructs across many languages and cultures.It is applicable across countries where the same language is spoken (e.g., French speakers in France and Switzerland) and across subcultures within the same country (e.g., Spanish speakers in New York and Los Angeles).Program materials (e.g., educational handouts and consent forms) can also be translated using this method.
Translating program materials and measures using the universal translation method involves several steps: (a) forward translation into the target group language by independent bilingual translators; (b) backward translation to compare the new document with the original document; (c) a review by the original document developer to ensure consistency between documents; (d) pilot testing the translated document with members of the target community to guide additional adaptations; and finally, (e) a second review by independent bilingual translators (Eremenco et al., 2005).After people familiar with the document's intent achieve translation consensus, it is proofread and reproduced.
Culturally appropriate intervention designs typically use strategies from several categories (Resnicow et al., 1999).Each intervention is unique, requiring a different set of strategies to achieve cultural relevance for the target population.For example, to achieve deep structure, socio-cultural adaptations may require incorporating constituent-involving strategies.To develop culturally equivalent measures, linguistic strategies may use strategies from the four other categories outlined above.For review purposes, the authors: 1) assessed the extent to which each of these adaptation strategies was present in each intervention study, and 2) how these strategies influenced the study results.

Inclusion and Exclusion Criteria
All childhood obesity intervention studies (e.g., diet, physical activity, parent-focused) since 2000 specifically targeting ethnic minority preschool children in the United States were included.Intervention studies not identifying the specific ethnic groups participating in the study were excluded.

Weighted, point-based scoring
A weighted, point-based scoring system was developed by the primary author to evaluate the overall depth and breadth of each intervention's cultural adaptation.Each of the five categories was assigned a weighted base score of 1 to 4 points.Less weight (1 point) was assigned to peripheral strategies involving surface structure adaptations reflecting cultural appearances (e.g., visual aids) and to evidential strategies addressing scientific evidence (e.g., risks for cardiovascular disease in obese children).More weight (2 points) was given to constituentinvolving and socio-cultural strategies addressing deep structure.Linguistic strategies received the most weight, based on the complex and time-consuming translation requirements for materials (2 points) and instruments (2 points).Finally, each category was assigned a weighted tailored score, based on the level of tailoring employed.The most weight was given to tailoring for individual participants (1 point), less to tailoring for a subgroup (0.67 points), and the least to tailoring for the ethnic group as a whole (0.33 points).Table 1 summarizes the scoring system.

Scoring procedures
A five-step procedure was used to analyze, score, and rank the interventions.Each intervention was first evaluated to determine which of the five cultural adaptation strategies were employed and whether the strategies were tailored.Next, each strategy category employed was analyzed and given a base score and a tailored score, which were then summed to create a total category score.All the category scores for each intervention were summed for a total adaptation score (15 maximum points).Finally, a normalized percent score was obtained by dividing the total adaptation score by 15, yielding the maximum possible score.
Prior to scoring, three interval ranks were selected: minimal adaptation (< 50%), moderate adaptation (> 50% and < 75%), and comprehensive adaptation (> 75%).The intervals were chosen so a moderately ranked intervention would employ at least two deep and one surface structure-based category, all tailored at the subgroup level.

Comprehensively Adapted Interventions
The two Fitzgibbon et al. studies (2005Fitzgibbon et al. studies ( , 2006) ) exhibited the highest cultural adaptation scores and comprehensive adaptation rank.Both used the same intervention design, targeting different ethnic groups.To tailor and culturally adapt their interventions, both studies used four of the five adaptation categories.For example, sociocultural strategies incorporated participants' requests for specific education materials (e.g., newsletters), class schedules, and safety considerations.Linguistic strategies for translating program materials and measures followed established guidelines.Furthermore, step-wise procedures were reported for the interventions' adaptations.(2007).No adaptations were reported for other instruments or program materials.

Minimally Adapted Interventions
The Williams and associates intervention (2004) required no oral or written input from participants.It simply directed the school foodservice to provide low saturated fat meals to the preschool children.No culturally adapted meals were provided.Educational materials were age appropriate, but no cultural adaptations to educational materials were reported.
The Worobey and associates' intervention (2004) also exhibited minimal adaptations.Only one of five adaptation categories was reported.A peripheral strategy addressing surface structure used bilingual public health nurses to facilitate the intervention.No cultural adaptations were reported for the intervention itself, materials, or instruments.
Cultural adaptations for the Harvey-Berino and Rouke ( 2003) intervention documented two out of five adaptation categories.Socio-cultural adaptations were incorporated by constituentinvolving community peer educators assigned to adapt the intervention.However, no specific adaptations were reported for the intervention, program materials, or instruments, resulting in a minimal ranking.
The McGarvey and associates' intervention (2004) also received a minimal rank.Cultural adaptations to the intervention included three out of five categories: peripheral, constituent involving, and linguistic.Spanish program materials were provided, but translation procedures for these materials were not described.

Intervention Outliers
By design, one of the interventions required minimal cultural adaptation.Thus, the intervention warranted no adaptation score.The Alhassan and colleagues' (2007) intervention was an observational study.No oral or written input was required from the children.Of note, this was the only intervention reporting to have culturally adapted the consent form.

Discussion
Major differences were identified between the comprehensively, moderately, and minimally adapted interventions.Interventions were characterized by: 1) the type and extent to which cultural adaptation strategies were present, 2) deficiencies in cultural adaptation, and 3) how adaptation strategies influenced their outcomes.

Type and Extent of Cultural Adaptations
Interventions ranked highest for cultural adaptation (Clarke, Freeland-Graves, Klohe-Lehman, & Bohman, 2007;Clarke, Freeland-Graves, Klohe-Lehman, Milani, et al., 2007;Fitzgibbon et al., 2005Fitzgibbon et al., , 2006;;Klohe-Lehman et al., 2007) incorporated four of the five strategy categories and involved surface and deep structure, as well as tailoring.The Fitzgibbon et al. studies (2005Fitzgibbon et al. studies ( , 2006) ) 2003) used a community-based participatory approach delegating responsibility for cultural adaptations to the Mohawk community.This is an effective strategy for incorporating the culture's surface and deep structure to design a relevant intervention (Horn et al., 2008), but the absence of documented adaptations and procedures precluded a higher rank.
Interestingly, none of the interventions used evidential strategies.Further research may be needed to evaluate the effectiveness of using this adaptation strategy with obesity interventions.

Deficiencies in Culturally Adapted Interventions
Notably, only 10 studies directed to ethnic minority preschool children qualified for review in the last decade.As displayed in Table 3, there were striking deficiencies in all the reviewed studies.Given the severity of childhood obesity, more research is needed on culturally adapted obesity interventions for high-risk preschool children.The paucity of such research studies is an example of the health disparities facing vulnerable populations (Branner et al, 2008;Brown et al, 2007: Bluford et al, 2007).
A major deficiency was the limited details about cultural adaptation and translation procedures for health promotion messages, materials, and measurement instruments.Elder, and colleagues (2009) reported that a majority of U. S. ethnic minority group members are functionally illiterate (reading at or below fourth grade) or marginally illiterate (reading between fifth and eighth grade).A common practice is to simply translate directly from English to the dominant ethnic language.This can render the translations culturally and linguistically inappropriate, especially for populations with low literacy levels.Appropriate linguistically adapted measures and program materials improve comprehension among low-literacy groups, which is essential to reduce reporting bias and ensure credibility of study results.Inappropriate or inadequate translation of measures could have introduced significant bias, threatening the integrity of the outcomes (Martinez et al., 2008).

How Adaptation Strategies Relate to Outcomes
The effect of cultural adaptation strategies on the success of an intervention is important.Assessing the success of an intervention, however, is neither simple nor straightforward.One issue is distinguishing the intervention shortcomings due to a faulty intervention design (e.g., inappropriate for its ethnic study group) versus a faulty intervention delivery (e.g., inappropriate translation of the educational materials) (Rychetnik, Frommer, Hawe, and Shiell, 2002).Intervention design as well as descriptive information provided by the author must be considered.Another issue is determining relevant dimensions of success to be considered.Evans (2003) proposed a hierarchy of evidence focusing on three dimensions for evaluation: effectiveness, appropriateness, and feasibility.Both the type of research design (e.g., randomized controlled trials, and descriptive studies) and the extent of cultural adaptations applied can influence these three dimensions.
Although it is beyond the scope of this review to perform a rigorous analysis of the 10 studies, some general conclusions can be drawn.It does appear the studies with higher adaptation scores (e.g., Fitzgibbon et al, 2005, andKlohe-Lehman et al, 2007), tended to be more appropriate for their ethnic groups, and demonstrated greater effectiveness and/or feasibility compared to the minimally adapted studies (e.g., Harvey-Berino and Rouke, 2003).Similarly, studies with moderate adaptation scores, (e.g., Clarke, Freeland-Graves, Klohe-Lehman, & Bohman, 2007;Clarke, Freeland-Graves, Klohe-Lehman, Milani, et al., 2007) tended to demonstrate moderate levels of effectiveness and feasibility.

Minimally ranked interventions in general
showed minimal effectiveness and feasibility.For example, in the Williams et al, (2004) intervention documented no cultural adaptations for educational materials.Study findings indicated no difference in outcomes between the intervention group receiving nutrition education and the intervention group without nutrition education.The lack of adaptation strategies may have influenced the outcomes, diminishing the intervention's effectiveness.This review has some limitations.The findings from these studies may be suspect for several reasons.First, lack of rigor in the study design can affect study outcomes.Most of the studies reviewed were not randomized controlled trials.Therefore, influences from confounding variables were difficult to identify or control.Second, inadequate cultural adaptation of the intervention may have resulted in non-equivalent measures.Strategies incorporating deep structure were also lacking and might have enhanced the relevance of interventions and outcomes for ethnic groups.
Based on the literature review search, no known scoring system for cultural adaptation was available.Therefore, a weighted scoring system was developed, as previously described.This scoring methodology is open to debate, as are the weights assigned to each adaptation category, the rank thresholds, and scores for each study reviewed.In addition, inadequate documentation of adaptation strategies used may have resulted in inaccurate scoring.Since few studies qualified for review, it is difficult to draw universally valid inferences regarding cultural adaptations.
Influences of adaptation strategies on study outcomes are very difficult to assess.In order for the effects of cultural adaptation on outcomes to be determined, there needs to be greater standardization in adaptation strategies used, with more attention to deep structure adaptations.Other factors that may be influencing outcomes, also need to be accounted for in study designs.

Conclusion
Standardized guidelines recognized and endorsed by health organizations, such as the Institute of Medicine and the American Nurses Association, are needed to improve the quality of culturally adapted interventions and translated program materials and instruments.Making standardized guidelines available to health promotion practitioners and investigators could help address health disparities and the paucity of effective interventions for high-risk vulnerable populations.Standardizing adaptation guidelines may be difficult, but will pay large dividends by enhancing the ability of investigators to efficiently design and tailor effective interventions for ethnically diverse populations.
Clearly, research interventions should be appropriately adapted and relevant for the target ethnic group.One suggestion is for funding agencies to require grantees to adequately document details of adaptation strategies used for interventions targeting ethnically diverse populations.This may improve intervention effectiveness and credibility of study results.It would also place emphasis on the importance of culturally adapting interventions and help reduce disparity in the quality of health promotion programs available to ethnic minority populations.
This review underscored the relative absence of culturally adapted obesity interventions for ethnic minority preschool children.Although not statistically rigorous, there seemed to be a consistent association between the extent of adapted interventions employed and the success of the outcomes.The most successful interventions in general were those that employed multiple cultural adaptations tailored to the intended ethnic populations.
Standardized cultural adaptation guidelines are needed for clinicians and investigators who are culturally adapting interventions.Findings suggest that documenting cultural adaptation strategies is crucial to support the integrity of study outcomes and permit study replication.There is also a need to examine, in depth, the influence of cultural adaptation strategies on intervention outcomes through carefully designed studies that account for other confounding variables.Finally, more research is needed for culturally appropriate interventions for ethnic minority preschool children at high risk for obesity.

*
Tailored Score: None = 0, Group = .33,Subgroup = 0.67, Individual = 1 V & R = valid and reliable ∑ = Sum Mat.= Materials Instr.= Instrument 2007; Three interventions exhibited moderate cultural adaptation.Klohe-Leman et al. (2007) used four of the five categorical strategies: (a) three peripheral visual aids for nutrition education, (b) one constituent-involving strategy, (c) four socio-cultural strategies integrating norms regarding ethnic foods, as well as economic and environmental factors, and (d) a linguistic strategy for one valid and reliable instrument.No adaptations were reported for the other instrument or program materials (e.g., handouts, curriculum).Both studies by Clarke and colleagues(Clarke,  Freeland-Graves, Klohe-Lehman, & Bohman,  2007;Clarke, Freeland-Graves, Klohe-Lehman, Milani, et al., 2007) adopted theKlohe-Leman et  al. (2007)  intervention design including cultural adaptations.The linguistic strategy for both interventions used the same valid and reliable instrument employed by Klohe-Leman et al.

Table 1 Scoring System for Cultural Adaptations
., Base Score = 2 if strategy used Lay health care workers, culturally sensitive staff Focus groups of target group members Bilingual/bicultural interviewers, educators, etc.