Tailoring lifestyle programmes for diabetes prevention for US South Asians

Mary Beth Weber , Monique M Hennink, K M Venkat Narayan

ABSTRACT

Key points

Tailoring Lifestyle Programs for Diabetes Prevention for US South Asians

Question:? The South Asian Health and Prevention Education pilot study used a mixed- methods study design to develop and test a culturally tailored, lifestyle intervention for overweight or obese South Asian Americans with pre- diabetes. This population develops diabetes more often, at younger ages and lower body mass indices, and has worse outcomes if they do develop diabetes than other race/ethnic groups. Finding effective and acceptable diabetes prevention programs is key to improving the health of this population.

Finding:

? South Asian Americans report that lifestyle choices around diet and physical activity are influenced by culturally prescribed family roles, the importance of South Asian foods in their lives, and changes due to immigration to the USA. In a pilot intervention study, participating in a culturally tailored lifestyle intervention programme modelled after the US Diabetes Prevention Programme resulted in improvements in cardiometabolic risk factors including weight, waist circumference, blood pressure, plasma lipids and glycated hemoglobin (HbA1c). Over half (55%) of participants’ glucose levels returned to the normal range at the end of the intervention.

Meaning:

? Diabetes prevention programmes can improve health markers in this high- risk population, but should include tools to increase and leverage social support, improve knowledge around physical activity, overcome cultural barriers and reflect cultural food preferences.

INTRODUCTION

South Asians, a rapidly growing and linguistically, religiously, and culturally diverse immigrant group in the USA,1have a uniquely high risk for type 2 diabetes, developing diabetes more often and at lower ages and body mass indices (BMIs) than other race/ethnic groups.23 Lifestyle interventions like the US Diabetes Prevention Programme (DPP), which promote weight loss through changes in diet, activity and behaviour, decrease type 2 diabetes incidence in high- risk individuals.4The DPP showed that lifestyle education was effective at preventing or delaying diabetes in all race/ethnic groups, including Asians,45 but the number of South Asians was small. Furthermore, there is limited research on how to best design and implement proven DPPs like the DPP among South Asian Americans. There is a need to develop and test diabetes prevention interventions targeted specifically to South Asian Americans. To increase programme uptake and retention, and therefore effectiveness, lifestyle programs need to be culturally appropriate and address the unique barriers to lifestyle change of South Asian Americans. Herein, the authors report the results of the South Asian Health and Prevention Education Study pilot study (SHAPE, clinicaltrials. gov #NCT01084928), which aimed to: (1) describe barriers to and motivators for health lifestyle behaviours among South Asian Americans and suggestions for diabetes prevention in this community through formative focus group discussions (FGDs) and (2) test the feasibility of a culturally tailored lifestyle intervention programme, modelled after the US DPP and modified based the data collected in aim 1, in a preintervention and postintervention study.

METHODS

The SHAPE study included: (1) formative FGDs; (2) intervention development based on the results of the FGDs and (3) a pre-post design trial testing the feasibility and effectiveness of the intervention. This mixed- methods study design ensured that the intervention design reflected the needs of the target community, increased acceptability of the intervention content and provided a richer evaluation of programme design and impact. Individuals provided written, informed consent before each FGD, at baseline testing, and before intervention enrolment. Deidentified data are available through the corresponding author.

A community advisory board including local leaders in the South Asian community provided guidance on recruitment and design. Participant recruitment for the FGDs and then the subsequent pilot study included advertisements in local South Asian magazines and on community organisation listservs. In- person outreach was conducted at events targeting South Asians (health fairs, diabetes education events, screening for an unrelated cardiovascular risk) and at meetings of South Asian clubs (for younger participants) and South Asian shopping centres and grocery stores (for older participants).

Formative qualitative research

The study team conducted FGDs to obtain a broad range of community beliefs6about lifestyle behaviours among South Asian Americans. FGDs were stratified by sex and age (25-39 or 40+ years) to create homogeneous groups of participants, minimise hierarchy development that may reduce data quality and enable structured analytical comparisons.6Trained moderators, matched to participants’ sex, used a semistructured interview guide to elicit conversation around lifestyle behaviours, barriers and motivators for exercise and healthy eating, and feedback on possible programme components. Moderators were not South Asian to encourage more detailed discussions, as would be needed when speaking to an outsider not familiar with their community.7The interview guide was not based on an existing framework, but was instead based on concepts from research literature around diet and physical activity behaviours and key components of the DPP. FGDs (n=17, n=109, average group size 6.4 individuals), five with older females and four each with younger males, younger females and older males) lasted 60-90 min, were digitally recorded, and were held at locations familiar to participants (eg, Indian restaurants, temples). Meaning saturation was reached around key themes central to the research goals (see the Results section) with no new information or nuance arising during FGDs.8All participants took a brief demographic survey before the discussion and received refreshments, parking validations and a gift bag. Participants were not contacted to provide further feedback after the completion of the FGDs.

The recordings were transcribed verbatim, deidentified and uploaded into MAXQDA (VERBI Software, V.10, V.11). The investigators developed codes based on a close read of the transcripts, and two analysts coded a subset of data independently to assess intercoder agreement. The researchers conducted a thematic analysis9of the data by developing thick descriptions of core influences on diet and physical activity behaviours and diabetes prevention in the South Asian community. Structured comparisons (men vs women, younger vs older participants) identified contextual differences between subgroups. The researchers calculated means and frequency distributions for survey variables.

Intervention design

The SHAPE curriculum, like the DPP, included 16 weekly core intervention classes, followed by maintenance classes (taught weekly in SHAPE instead of monthly like the DPP), mirrored the topics covered in DPP lessons and encouraged participants to reach weight (7% wt loss) and physical activity goals (150+ min per week). Modifications to the DPP curriculum based on the FGD analysis and feedback from the community advisory board are described in the Results section.

Pilot intervention

South Asians aged ≥25 years with BMI ＞22 kg/m2(indicative of overweight/obesity in Asians10and pre- diabetes (fasting glucose 100-125 mg/dL (impaired fasting glucose, IFG) and/or 2 hours postload glucose of 140-199 mg/dL (impaired glucose tolerance, IGT)11were eligible for the pilot study. Individuals with diagnosed type 1 or type 2 diabetes or taking diabetes medication, currently pregnant or with a history of heart disease, serious illness or conditions that may impede or prohibit participation were excluded. Prescreening (n=117) was conducted in- person or over the phone and included demographic questions and the FINDRISC (Finnish Diabetes Risk Score) screening tool, a validated diabetes risk score,12and at in person screenings, measures of height, weight and waist circumference. Individuals with a FINDRISC score of ≥11 were invited for baseline testing.

Baseline (n=55) and follow- up testing at months 4 and 6 were conducted at the Georgia Clinical and Translational Science Alliance Emory University Hospital Clinical Interaction Site. Testing included fasting blood draws to measure plasma lipids, insulin, glucose, A1C, complete blood count and blood chemistry; anthropometric measurements (weight, height, percentage body fat, waist and hip circumference, abdominal sagittal circumference and full body bioimpedance); blood pressure measurement; study questionnaires; a medical assessment and a 24- hour dietary recall. A 75 g oral glucose tolerance test with blood draws at 0, 30 and 120 min was done at baseline and month 6. Eligible participants (n=17) completed an exercise treadmill test13to identify contraindications to unsupervised exercise; none were found.

Lifestyle classes of 3-8 participants were held at two locations across Atlanta on weekend days or weeknights. Participants selected their preferred class. Classes lasted about an hour, included a weight measurement and were led by a trained health educator.

Quantitative analysis was conducted in SAS, V.9.3 (SAS Institute). Descriptive analysis was conducted for all variables and baseline and 6 months values were compared using t- tests (for continuous variables) or X2tests (for discrete variables).

RESULTS

Focus group discussions

Focus group participants (n=109, table 1) were from India, Pakistan and Bangladesh and were primarily Hindu, but groups also included Muslims (22.9%), Christians (4.6%), Sikhs (3.7%) and Jains (3.7%). The top five languages spoken at home were Gujarati (40.4%), English (33.9%), Hindi (25.7%), Tamil (8.3%), and Urdu and Punjabi (6.4% each).

Barriers to a healthy lifestyle

Participants cited that unhealthy lifestyle behaviours in their community, including physical inactivity, diets high in fat and refined carbohydrates, large serving sizes and stress, where common. Immigration, strongly defined gender roles and the cultural importance of food (table 2) influenced diet and exercise choices. After immigrating to the USA, the abundance of inexpensive food, reliance on convenience food products and the profusion of cars and elevators made it easier for participants to eat less healthy diets and be more sedentary. Migration was especially detrimental to the diet of South Asian men because they lacked cooking skills and therefore relied on preprepared, frozen foods or restaurant meals, food choices that are often high in fat, sodium and calories.

Culturally prescribed gender roles also affect lifestyle behaviours. Men are expected to be ‘providers’ and women the ‘carers,’ with both prioritising their family’s needs above their own health. Male participants said they are expected to succeed academically, excel at work and save money, leaving little time for exercise. This emphasis on material success also manifested as valuing saving money now over saving health and healthcare dollars later; one older male said: ‘You’ll pay tons of dollars to the doctor, but if you say, here, to buy a gallon of milk it’ll cost you two dollars extra every day, you aren’t willing to do that and so we end up paying much higher value later on.’ South Asian culture dictates that women prioritise caring for their family and household over their individual needs and taking time to exercise for health, pleasure or other reasons is discouraged.

Participants described the integral role of food in South Asian culture, family life, socialising and celebrating. Although women are responsible for food preparation, they find it difficult to improve their families’ diet. Taste drives food choice, and ‘oil is the king… to make [food] taste, uh, delicious’ (Young Male). Participants in each group felt ‘healthy’ and ‘delicious’ were opposites, and people, especially men, felt that lowering the fat in South Asian food would make it less, or even unpalatable, palatable. Women report being called a ‘scrooge’ if they use less oil in cooking, particularly at social occasions, and attempts to lower fat or sodium were often met with resistance.

Age-gender group＊# of groups Hindu, %Married, %Mean years in the USA US born, %Home ownership, %History of diabetes?, %Education IeveI Older males 4 62.5 94.3 0 22.1 65.7 55.6 ≥Primary school Older females 5 69.2 72.0 0 25.1 77.0 48.5 ≥Primary school Younger males 4 67.8 42.8 24.3 10.6 48.0 7.0 ≥College Younger females 4 80.3 37.8 29.0 11.8 72.3 0 ≥College

Overcoming barriers to healthy lifestyle behaviours

South Asians who make and maintain healthy lifestyle choices credit the support of friends and family. They sought out activities with social aspects (eg, group sports for young men, family walks for older participants, or exercise classes for women) to help them maintain motivation and/or had friends and family who encouraged them verbally or by example. Family was a motivator for and not a barrier to lifestyle improvement for healthy lifestyle adapters. They decided to exercise and follow a healthy diet to better care for family and prevent disease and disability to have more quality time with their loved ones. These participants were also motivated to disseminate healthy lifestyle knowledge to others in their family network. For example, one young man shared: ‘When I came [to the US], and when I educated myself, I brought this rule [for when my parents visited the US]: Okay, you have to eat this in the breakfast, eat this thing in the lunch, eat this, eat this thing in the dinner. And, after when they went back [to India], they’re still following it. So, if somebody is really strong in a family, they can probably bring this change.’ Women overcame their family’s reluctance to diet change by gradually altering foods and recipes (eg, slowing increasing spice while decreasing oil) without their family’s knowledge. These women reported that their families never noticed the changes and now enjoy healthier foods.

Suggestions for lifestyle interventions in the South Asian community

In all groups, participants felt strongly that an intervention like this was needed to prevent the spread of diabetes in the South Asian community. While most participants believed the intervention would be important for all age groups, younger men often discussed the intervention in terms of its importance for older South Asians.

In terms of class logistics, most participants agreed that English language classes would be preferable, reporting that most US South Asians, except for a few older individuals, spoke English: ‘Everybody knows English. We will say 99.9 percent know English and do understand what you are talking about’ (Older Male). Online content could be used to support in- person classes, but the majority of participants felt in- person classes would be most effective because they build peer support and create a culture of accountability. In particular, participants felt that in- person exercise instruction was required; ‘I think- I- there’s always the social aspect of it, too. I mean, I’m not talking to anybody necessarily, but when you’re around people, it kinda motivates you’ (Younger male). All groups reported that classes should be near participants’ homes at familiar locations. Young men favoured worksite- based programs, and women in both age groups stated that childcare should be provided. Class instructors could be ‘a(chǎn)nybody’ as long as they were ‘the best qualified person,’ although some female participants suggested that a South Asian dietician might be needed to provide dietary guidance.

In terms of content, participants in all groups reported a need for nutrition information for South Asian foods. Younger participants requested that the class include information on foods from a variety of cuisines ‘because our generation will eat everything’ (Younger Woman). Several participants stressed that there should be specific dietary advice for vegetarians and non- vegetarians, as it was felt these diet patterns differed significantly. Although women reported an interest in cooking demonstrations and recipe books, these ideas were less appealing to the male groups. Younger men and women described the need for quick cooking Indian recipes and cooking techniques. Women reported a desire for exercise classes that taught a variety of exercise types, for example, yoga, aerobics and even sports, which they felt would appeal to younger South Asian men. Except for older women, all groups reported a need to tailor exercise classes specifically based on the participant characteristics (eg, older adults or families).

Intervention design

Based on the results of the FGD and community advisory board feedback, changes were made to the intervention protocol and the DPP curriculum. A planned control group was removed from the study design as it was deemed unnecessary given the strong evidence for the D PP and advisement that it would affect recruitment and community interest in the study. Modifications from the DPP are shown in table 3. These include adapting dietary examples and advise to be more culturally relevant, enabling social support for lifestyle changes, encouraging participants to build support for healthy changes and providing a strong foundation for exercising. The SHAPE curriculum was based on the lesson plans used in the original DPP study;4since that time, some of these changes have also been independently added to the National DPP (eg, classes are now all group based).

Pilot intervention trial results

Of the 55 people screened, 22 were eligible for the SHAPE pilot study ( clinicaltrials. gov NCT01084928) and 17 consented to participate. Participants and non- participants were similar except participants felt they were at low risk for a serious disease (52.9% vs 28.9%, p=0.037) and presented more often with isolated IFG (iIFG, 41.2% vs 2.6%, p=0.0002) or IFG+IGT (35.3% vs 5.3%, p=0.0035). Pilot participants were 76.5% male with a mean age of 46.9±12 years and had some college or higher education. Mean BMI was 28.9+/-6.1 kg/m2, and 41.2% had isolated IGT, 17.7% had iIFG and 35.3% had IFG+IGT (35.3%). One participant with normoglycaemia was included in the classes but not in programme assessments at his request and because he could provide valuable input on the feasibility of the programme among young men.

Nine pilot intervention participants returned for 6- month follow- up testing. Two participants dropped out of the pilot, one for health reasons not related to the study and the other citing a lack of interest in continuing the programme. Of the remaining participants who declined attending follow- up testing visits (n=4), they cited the inconvenience of travelling to and parking at Emory University and difficulty taking time off work as reasons. Eleven participants attended more than half of the active intervention period classes, and eight participants attended at least half of the maintenance period classes.

As shown in table 4, five of the nine study participants attending 6- month testing had reverted to normoglycaemia. Mean weight, BMI, waist circumference and abdominal sagittal circumference all decreased, although none of these improvements reached statistical significance. Participants also had a slight decrease in systolic blood pressure and showed marked improvements in cholesterol, low- density lipoprotein and triglycerides. Results are similar when the analysis is restricted only to individuals with 6- month follow- up data (data not shown).

DISCUSSION

This study used mixed methods to plan, implement and test the feasibility of a lifestyle intervention programme for diabetes prevention for US South Asians. Formative qualitative research found that South Asian Americans report a high prevalence of unhealthy behaviours and felt that immigration, culturally prescribed gender roles and the role of food in their culture were significant barriers to making healthy lifestyle choices. Some participants used many of these same factors-particularly the importance of taking care of their family and the need to be healthy to fulfil this role-as motivating factors for making healthy changes. In a subsequent pilot intervention study, a lifestyle education programme based on the DPP successfully recruited and enrolled South Asian Americans and led to positive improvements in glycaemic status, anthropometry and plasma lipids. However, on- campus testing was a barrier to attending follow- up visits.

“踩爆它！”有學(xué)生立馬回答。我笑著說：“可以，不過要找一個(gè)安全的地方，不要傷害到自己，也不要嚇到同學(xué)。想怎么處理你們裝滿壓力的氣球，是你們的權(quán)利，也是你們的智慧！”

?

The SHAPE study supports research from UK, Canada and US South Asians describing the important role of food in South Asian culture,14the prevalence of unhealthy dietary behaviours,15-20and how migration changes lifestyle behaviours, resulting in weight gain and increased risk of diabetes.21-23However, the ways that migration effects behaviour differs for South Asians in non- US settings; for example, in a study of South Asians in Australia, participants blamed poor eating habits and a move away from traditional home- cooked foods on underemployment, the necessity of dual- income households and eating cheaper, less healthy food options because of financial difficulties.24Conversely, SHAPE participants discussed using fast foods because of the lack of cooking skills (for men) or because of limited vegetarian choices and an increase in potentially unhealthy ingredients (eg, sugar, oil) because they lower cost in the USA allowed participants to splurge.

South Asian society is pluralist,25-27prioritising the needs of the community over self; it is perhaps unsurprising that participants reported that family responsibilities are the major barrier to increasing physical activity. Although SHAPE participants cited many of the same barriers to physical activity as other South Asian migrant studies (household or work responsibilities, little cultural focus on being active), SHAPE participants did not mention current health conditions as barriers to physical activity as has been reported elsewhere.28-31

Some of the differences in reported barriers and activities around lifestyle behaviours in the SHAPE population may be due to the relatively high education and socioeconomic status of participants. SHAPE focus group participants understood of the link between healthy diet, physical activity and chronic disease prevention, even if this knowledge did not translate into changed behaviours. In other studies, South Asian participants did not make the connection between lifestyle choices and disease risk or current disease status.242932

DPPs for South Asians can use the same cultural influences that currently pose barriers to healthy lifestyle choices as forces of change. Participants in this and other studies2428 30 33reported that social support, particularly from family, was vital for starting and maintaining diet change and physical activity. Family was also a primary motivating factor for healthy lifestyle changes, and families that changed their lifestyles together were particularly successful.

VariabIe BaseIine for participants with foIIow- up data(n=9)BaseIine for aII participants(n=17)Six months(n=9)GIycaemic status Normoglycaemia 1 (5.9%) 0 (0%) 5 (55.6%) Isolated IGT 3 (17.7%) 3 (33.3%) 2 (22.2%) Isolated IFG 7 (41.2%) 3 (33.3%) 1 (11.1%) IGT+IFG 6 (35.3%) 3 (33.3%) 1 (11.1%) Type 2 diabetes mellitus 0 0 0 HbA1c (%)＊ 5.8 (0.4) 5.9 (0.5) 5.7 (0.4)Anthropometry Normal weight BMI 1 (5.9%) 0 (0%) 3 (33.3%) Overweight BMI 7 (41.2%) 2 (22.2%) 2 (22.2%) Obese BMI 9 (52.9%) 7 (77.8%) 4 (44.4%) High waist circumference 14 (82.4%) 8 (88.9%) 5 (55.6%) BMI (kg/m2) 28.9 (6.1) 30.7 (5.1) 26.2 (3.9) Weight (kg) 78.1 (13.8) 80.3 (11.2) 73.9 (14.8) Waist circumference (cm) 97.9 (10.9) 99.7 (8.9) 92.7 (12.0) Abdominal sagittal circumference (cm) 23.6 (6.4) 25.6 (7.6) 20.9 (5.2)Other cardiometaboIic markers Systolic blood pressure (mm Hg) 121.8 (9.8) 124.3 (9.4) 118.3 (14.2) Diastolic blood pressure (mm Hg) 78.8 (8.6) 78.3 (8.8) 77.1 (9.4) Total cholesterol (mmol/l) 4.8 (0.9) 4.9 (1.0) 4.4 (0.7) LDL (mmol/L) 3.3 (0.6) 3.4 (0.8) 3.0 (0.7) HDL (mmol/L) 1.0 (0.2) 1.1 (0.3) 1.0 (0.3) Triglycerides (mmol/L) 1.7 (0.7) 1.5 (0.5) 1.2 (0.5)

There is limited research applying these and similar findings for diabetes prevention in South Asians. Outside of the USA, studies delivering the DPP or the Finish Diabetes Prevention Study curriculum have reported mixed, but largely positive impacts on glycaemic status with little effects on anthropometry.34-36Two US studies of lifestyle programmes for reducing diabetes or cardiovascular risk reported improvements in glycaemic markers and in one study weight.3738

Although the SHAPE pilot study was small and lacked good follow- up for study testing, it has promising results that support further testing of culturally tailored lifestyle intervention programs for the US South Asian population. Follow- up in study classes was good, with only two participants dropping out of the programme, and those who attended follow- up visits showed positive improvements in anthropometric markers such as body weight and more markedly, waist circumference. In the DPP, weight loss was the strongest predictor of diabetes risk reduction.39Perhaps even more striking is that five of the nine participants at end of study testing had regressed to normoglycaemia; within the DPP, even transient regression to normal glucose levels was associated with a 56% reduction in diabetes risk.40

This study has several strengths. The formative and pilot trial data add to the small but growing literature on diabetes prevention and lifestyle behaviours among US South Asians. The FGDs included a large and diverse population of South Asian adults, including men and women with and without diabetes from various countries and regions of South Asia, of different religious traditions, and with a range of education levels. The results varied little between age and sex groups indicating that the barriers and motivating factors for diet and physical activity change are not patterned by standard demographic variables, but rather are stronger cultural issues seen in a diverse population of South Asian Americans. The pilot trial adds important information on the feasibility of translating proven DPPs to the South Asian Americas. This was the first study to our knowledge to report regression to normoglycaemia among South Asians enrolled in a diabetes prevention study, and SHAPE is one of the few studies of diabetes prevention among South Asians, enrolled participants across the pre- diabetes spectrum, and specifically targeted overweight and obese South Asians.

Nevertheless, this study has several limitations. The pilot study too small to enable significance testing of results and there was loss of participants at follow- up testing visits. The goal of the SHAPE intervention was to assess programme feasibility, however, and even with these significant hurdles, the study provided a better understanding of barriers to participation and possible effects of the programme on health outcomes. Recruitment logistics for FGDs made it impossible to recruit younger participants quickly enough to host separate groups for first and second- generation South Asians resulting in an inability to capture differences that place of birth might have on lifestyle behaviours. The results of this study may not apply to non- English- speaking South Asians. The researchers believe that this is not the case, since focus group participants described behaviours of the South Asian community more broadly and requested the programme in English, and the research team was able to recruit a sizeable population of older South Asians where English skills might be lowest. However, acculturation affects lifestyle behaviours such as diet and physical activity, and the researchers cannot rule out the possibility that English- speaking participants are more acculturated than other members of the South Asian community and that different interventions might be needed for South Asians with low- English proficiency.

Despite these limitations, the SHAPE study provides important information on the current behaviours, beliefs and feelings around diabetes prevention, diet change and physical activity, as well as pilot data showing the feasibility of a DPP- like intervention among South Asian Americans. Future work should focus on testing the SHAPE intervention in a larger trial, collecting more formalised feedback on programme acceptability and considering additional intervention models, a family- based programme, for instance, to address the high risk of diabetes in this community.

CONCLUSION

In the USA, where food is plentiful, portions are large, unhealthy ingredients are cheap and readily available, and the environment supports sedentary activity over active lifestyle choices, South Asians, with their focus on family, society and food, are at a uniquely high risk for weight gain and developing cardiometabolic diseases like diabetes. This study showed that, although immigration and traditional gender roles can act as barriers to following a healthy lifestyle, they could also motivate South Asians to change their diet and physical activity behaviours. Lifestyle interventions, based on the US DPP, can be an effective tool for reducing weight, improving glycaemia and engaging South Asian adults in diabetes prevention.

ContributorsMBW designed and managed the SHAPE study, conducted data analysis and was the primary author of the manuscript. MMH provided guidance on the development and implementation of the qualitative components of SHAPE. All authors contributed to the design of the SHAPE study, assisted in data interpretation and provided substantive feedback on the manuscript.

FundingAs requested, we deidentified the manuscript. The information removed was Emory as the institution in several places (currently 'UNIVERSITY') and the initials and grant information in the funding statement. The full statement is: This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (NIDDK/NIH) (grant number R34 DK081723), the Clinical and Translational Science Award programme, NIH, National Center for Research Resources (grant number UL1 RR025008), the Emory University Coalition of University- Community Partnerships, and the American Diabetes Association. Further support was provided for MBW by additional grants from the NIDDK/NIH (grant numbers T32 DK007298, T32 DK007734) and both MBW and KMVN receive support from a NIDDK/NIH- funded Centre grant (grant number and P30 DK111024).

Competing interestsNone declared.

Patient consent for publicationNot required.

Ethics approvalThe Emory University Institutional Review Board approved the FGDs (#IRB00019630) and the intervention (#IRB00035893).

Provenance and peer reviewNot commissioned; externally peer reviewed.

Data availability statementData are available on reasonable request. Deidentified data are available through the corresponding author.

Open accessThis is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY- NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non- commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non- commercial. See: http:// creativecommons. org/ licenses/ by- nc/ 4. 0/.

ORCID iD

Mary Beth Weber http:// orcid. org/ 0000- 0002- 2804- 3948