Background
Rates of hyperglycaemia in pregnancy – due to pre-existing type 2 diabetes or gestational diabetes (GDM) – are disproportionately higher among Indigenous women worldwide [
1,
2]. In the Australian context, Aboriginal women are 1.2 times more likely to have GDM and greater than 10 times more likely to have pre-existing type 2 diabetes than non-Indigenous women [
3]. Recurrence of GDM is reported to occur in 30–84 % of subsequent pregnancies [
4] and women with GDM have a sevenfold higher risk of developing type 2 diabetes compared to women without GDM [
5]. The risk is higher for Aboriginal women who have a greater than fourfold risk of developing type 2 diabetes after GDM, compared with non-Indigenous women [
6].
Evidence suggests that type 2 diabetes can be prevented with lifestyle and pharmacological methods [
7], although the evidence in the post GDM population is not as clear [
8,
9]. Trials showing type 2 diabetes can be prevented with lifestyle change, such as the Diabetes Prevention Program, generally involve expensive and time-consuming interventions [
10]. This presents a challenge for mothers with young families, who commonly cite tiredness, and competing work and carer duties as barriers to lifestyle change in the postpartum period [
11,
12]. To date, this evidence is primarily based on research of non-Indigenous women in urban centres and may not be applicable in other contexts. Whilst there are several lifestyle programs targeted at Indigenous populations [
13,
14], there is currently no literature on specific postpartum programs in Indigenous populations worldwide.
In order to be successful, lifestyle programs need to be context specific and adaptable, ensuring they are culturally appropriate [
15]. Aboriginal Australians are culturally and linguistically diverse peoples who now comprise 3 % of the Australian population. In the Northern Territory (NT), 30 % of women identify as Aboriginal and/or Torres Strait Islander and 80 % of those women live in remote communities [
16]. Aboriginal women in remote Australia, experience high levels of social disadvantage, including inadequate housing, food insecurity [
17] and inadequate places to exercise, serving as barriers to engagement in lifestyle change programs. Conversely, Aboriginal cultures have strong connections to ‘Country’ (a term that refers to traditional lands [
18]) a potential enabler to participation in lifestyle programs.
In light of perceived barriers to lifestyle programs, pharmacological interventions have been evaluated. Metformin has been shown to prevent type 2 diabetes in the general population [
19], and to be more effective at preventing type 2 diabetes in women who had GDM compared to parous women without a history of GDM [
20,
21]. However, there are concerns around adherence to metformin given its gastrointestinal side effects, and the large size of the tablet [
22]. Whether metformin used for primary prevention in the Aboriginal Australian context would be an acceptable and effective strategy is currently unknown.
Another important consideration for lifestyle interventions is that concepts of health for Aboriginal populations differ from the Western biomedical paradigm that drives mainstream healthcare delivery. Programs need to privilege Aboriginal people’s voices and consider the different ways cultures conceptualise health [
23]. Furthermore, individual level programs are unlikely to be successful without addressing the underlying societal and economic drivers at play [
24,
25]. We aimed to explore Aboriginal women’s and health providers’ preferences for a program to prevent and improve diabetes after pregnancy.
Methods
This qualitative study was underpinned by a phenomenological methodology [
26]. Interviews were conducted, with a focus on how perceptions of a diabetes prevention program were shaped by the experience of diabetes in pregnancy. It also aimed to empower and incorporate Aboriginal voices and cultural values within this research [
27,
28]. This study sits within The Diabetes across the Lifecourse: Northern Australia Partnership, a partnership between researchers, policy makers and health service providers. Strategic advice concerning aspects of the Partnership is provided by an Aboriginal and Torres Strait Islander AdvisoryGroup (ATSIA), including Aboriginal women across the NT and represented here by SG, chair of the ATSIAG. The design of this study was in response to a priority set by the ATSIAG to prevent type 2 diabetes after GDM. Prior to study commencement, the ATSIAG. emphasised the importance of employing an Aboriginal Community Worker and returning to communities to ensure outcomes from this project were communicated back to those involved. In each community an Aboriginal Community Worker was recruited through established relationships between the Partnership and the communities. Whilst they had no formal research training, both women had experience in cross-cultural communication and fluency in both English and local languages. They were employed by the research team to establish a culturally safe space and ensure the appropriate interpretation of Aboriginal women’s voices.
The research team comprised of Aboriginal and non-Indigenous people, vital to establishing rapport, collecting meaningful data and making accurate interpretations of data. AW and BMR were responsible for conducting interviews and AW and SG were responsible for analysing the data. AW is an endocrinologist and PhD candidate of European ancestry. The potential power imbalance from her position was, at least in part, mitigated by interviewing in partnership with an Aboriginal community worker. BMR is an Aboriginal Wiradjuri woman with a masters by research and interviewer experience. SG is an Aboriginal Noongar and Bardi-Jawi woman, with a certificate IV in Indigenous Research Capacity Building and a Bachelor of Applied Science. RK is of European ancestry, and an early career researcher with expertise in qualitative methodologies and health systems research. RK led supervision of data collection, analysis and interpretation.
Study setting
Participating communities responded to an initiative requesting communities to undertake this study with members of their staff and clients from Top End Health Service Remote Primary Health Care clinics in the NT. These communities (A and B) are situated in the NT and classified as very remote and socio-economically disadvantaged [
29]. Local Aboriginal languages are spoken in each community with English being the 3rd or 4th language spoken.
Study participants
The study population were Aboriginal women with a history of GDM or type 2 diabetes in pregnancy in the last five years, aged > 18 years and purposefully sampled based on these criteria. Health professionals at the remote clinics identified potential participants and the Aboriginal Community Worker in each community informed potential participants of the study and assisted with gaining consent. Health professionals were recruited through direct contact with remote clinics. A group of community advocates who self-identified as community elders were also interviewed (4 women together). This study was approved by the Human Research Ethics Committee of the NT Department of Health and Menzies School of Health Research (HREC 19-3362).
Data collection
The content of the interview questions was developed and informed from a literature review and input from the research team, then discussed with the ATSIAG and the Aboriginal Community Workers, with no concerns raised. The interviews focussed on participants’ opinions on lifestyle changes and metformin for primary prevention of diabetes. The interview guide was piloted with initial participants (included in data analysis) and modified over the study to ensure content relevant to the aims of the study was captured and framed in an appropriate way. No further interviews were conducted once consistent findings emerged from interviews.
AW and BMR conducted interviews in community A in the presence of the Aboriginal Community Worker and AW conducted interviews in community B in the presence of an Aboriginal Community Worker from October 2019 -February 2020. Interviews were conducted in English and questions and responses in local language were translated in English by the Aboriginal Community Worker when needed. Interviews occurred at a location deemed appropriate by the participant and Community Worker (health clinic, participants’ homes or outside in a private space). Interview duration ranged from 20 to 60 min and field notes were taken by AW.
Data analysis
Interviews were audio-recorded and transcribed verbatim by an independent transcription company. Immediately after each interview (with Aboriginal women), interpretation of meaning was discussed between AW and the Aboriginal Community Worker to ensure AW correctly interpreted women’s data and field notes were taken by AW. Participants were invited to review their transcripts, though none did.
The program NVivo (version 10) was used to guide analysis. AW and SG independently undertook the initial round of coding -analysing both transcripts and field notes- using an inductive analysis framework [
30]. This involved line by line coding whereby clusters of meaning informed the development of categories highlighting participants' descriptions of experiences. Evolving themes were identified and negotiated between AW and SG until a final coding structure was determined. To cross-check the accuracy of interpretation, after the initial rounds of coding, it was planned that the Aboriginal Community Worker in each community would review themes developed. However, due to the COVID-19 pandemic, remote communities were closed for non-essential visitors in 2020 and AW was unable to revise themes in person with the Aboriginal Community Workers. Instead, AW discussed key themes with the community worker from community B via telephone, although was unable to contact the Aboriginal Community Worker in community A. Additionally, key themes were discussed with the ATSIA for advice, feedback and guidance. Recommendations from the study will be shared in feedback sessions with communities once COVID-19 pandemic restrictions are lifted.
Discussion
In this study, we engaged health providers and Aboriginal women with the aim of exploring their views on a postpartum program. We report four key findings. Firstly, individual and community level programs must occur in parallel with structural changes designed to address food insecurity and adequate facilities to exercise. Secondly, programs need to be community driven and co-designed with participants. Thirdly, programs should be grounded in Aboriginal conceptions of health and finally, lifestyle programs were preferred over metformin.
Community lifestyle interventions should occur in parallel with policies at a government level. Like other remote Aboriginal communities, community A and B have experienced a recent history of colonisation with enduring effects, including substandard housing, food insecurity, psycho-social stressors and a lack of economic and occupational opportunities [
31]. Although traditional food practises continue to play an important role within community, colonisation brought rapid lifestyle change with adverse changes in physical activity and nutrition [
32,
33]. In our study, food insecurity was a key issue repeatedly identified by participants as a barrier to eating well. Gorton et al’s review on interventions to address food insecurity in high-income countries, reported a lack of evidence that community programs alone effectively reduce food insecurity [
34]. Instead, governmental efforts to improve household incomes and reduce overcrowding, alongside community programs, would likely go further to reduce food insecurity [
34].
There is substantial evidence for the benefits of healthy diet and physical activity in preventing and improving type 2 diabetes and cardiovascular disease [
5]. However, despite the high rates of these conditions among Indigenous populations, examples of lifestyle programs are rare. Two systematic reviews on physical activity in Indigenous populations in North America [
35] and in Australia and New Zealand [
36] report both a small number of interventions and a lack of evaluations of these interventions. Hence, it is currently unclear how successful programs are, at increasing activity levels for Indigenous populations worldwide, and certainly there is no evidence for programs in the postpartum period.
What is clear, and consistent with our findings, is that programs initiated by the community are more likely to be sustained and have positive health effects [
15,
37,
38]. A systematic review of nutrition programs for Aboriginal Australians report the most important factor determining the success of such programs is community control of development and implementation [
15]. The Healthy Communities Project was a multi-component strategy that partnered with community leaders to reduce sugary drink consumption in remote communities in Queensland. Engagement from a range of stakeholder groups resulted in increased water sales and decreased sugary drink sales [
39]. A community driven diabetes prevention program in a Maori rural community in New Zealand had similar positive results, with the investigators attributing its success to community ownership [
40]. Our research, with participants reporting a desire for programs to be community driven, adds to the body of literature supporting the co-design of lifestyle programs. Co-design is considered best practice in research involving Indigenous peoples, and defined as meaningful end-user engagement in research design with engagement across all stages of the research process, with clear guidelines in New Zealand, Australia and Canada [
41‐
43]. As evidenced by a diabetes prevention program for Aboriginal people in Western Australia, direction from community was vital for ensuring relevant application of previous research and effectiveness [
44]. .
Connection to and appreciation of Country holds great importance to many Aboriginal People [
45]. As reported by Thompson et al., being 'on Country' plays an important economic, dietary and cultural role in many communities [
45] and we found participants expressed improved physical and mental health when they felt connected to Country and expressed a desire for dietary programs to incorporate traditional food practices. Similarly, in a study among Cree women with a history of GDM, the importance of incorporating traditional elements into a diabetes prevention program was underscored [
46]. A systematic review on the benefits associated with engagement in Aboriginal Land Management activities (Australia and Canada), reported such engagement to be associated with significant health benefits through improvements in diet, activity, autonomy, and social and spiritual connection to land [
47]. Like exercise programs, rather than approaching diet through the biomedical model, Aboriginal peoples’ understandings of food systems and well-being should provide the foundation from which to reset the narrative in relation to diet and health [
48].
One’s primary role as a family and community member was evident throughout our interviews. As described in the literature [
37], many of our participants expressed shame if participating in exercise for personal reasons rather than for the benefit of the community or family. Whilst the concept of shame has been variously described, in Aboriginal culture it can refer to situations in which a person is singled out, or the centre of attention and the associated concern regarding what others would think [
49]. Shame is reportedly one of the largest hindrances to young Aboriginal women participating in physical activity [
50]. Hence, individual style programs such as walking groups are unlikely to work in this context. Generally, women in our study were not interested in joining such groups due to (i) the individual nature and associated shame and (ii) the reliance of previous programs on people outside of the community to run them and the transient nature of such people. Alternatively, group sports, that give women a sense of collective identity from being part of a team, are more likely to be effective and could aid in keeping women connected to family and community [
51]. In addition, women expressed a desire for group sport to be for women only.
Given the known barriers to lifestyle change, we asked participants for their perspective on using metformin to prevent diabetes. Metformin was not preferred by most of our participants. There were concerns with adherence given its gastrointestinal side effects and large size and practical considerations around length of treatment. Furthermore, all participants were able to identify clear gaps in lifestyle optimisation and expressed a preference to address these gaps prior to medicalising what many considered to be an environmental and structural issue.
Strengths of this study include the extensive involvement of (i) local Aboriginal Community Workers who were crucial in establishing a culturally safe space and ensuring appropriate interpretation of Aboriginal women’s voices and (ii) the guidance of the ATSIAG, who were actively involved throughout the project. There were several limitations. Generalizability and replicability of our findings are limited. Indigenous women are diverse, and any new program needs to be adapted to meet the specific needs of women and their cultural context. We were unable to return to communities to discuss themes in person with Aboriginal Community Workers due to COVID-19 travel restrictions. However, to cross-check the interpretation of findings and minimise biases, AW and the Community Worker spoke after each interview and SG analysed all transcripts. There was a smaller number of women recruited from community B compared to community A due to travel restrictions; however, themes were consistent across the two communities.
Acknowledgements
The authors would like to acknowledge the participants of the study, The Diabetes across the Lifecourse Northern Australia Partnership investigators, partners, staff, Clinical Reference Group, Aboriginal and Torres Strait Islander Advisory Group, NT health professionals from NT Department of Health hospitals and remote primary health care who have supported this work. Investigators of The Partnership’s Diabetes in Pregnancy Program, in addition to those named as authors are: E Barr, F Barzi, A Brown, S Campbell, S Corpus, B Davis, A Hanley, R McDermott, A McLean, J Mein, E Moore, J Oats, K O’Dea, A Sinha, J Shaw, M Wenitong and P Zimmet.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.